Classifications

• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T7/00—Image analysis
  • G06T7/0002—Inspection of images, e.g. flaw detection
  • G06T7/0012—Biomedical image inspection
• • A—HUMAN NECESSITIES
  • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
  • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
  • A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
  • A61B6/02—Arrangements for diagnosis sequentially in different planes; Stereoscopic radiation diagnosis
  • A61B6/03—Computed tomography [CT]
  • A61B6/037—Emission tomography
• • A—HUMAN NECESSITIES
  • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
  • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
  • A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
  • A61B6/50—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment specially adapted for specific body parts; specially adapted for specific clinical applications
  • A61B6/507—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment specially adapted for specific body parts; specially adapted for specific clinical applications for determination of haemodynamic parameters, e.g. perfusion CT
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T2200/00—Indexing scheme for image data processing or generation, in general
  • G06T2200/24—Indexing scheme for image data processing or generation, in general involving graphical user interfaces [GUIs]
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T2207/00—Indexing scheme for image analysis or image enhancement
  • G06T2207/10—Image acquisition modality
  • G06T2207/10072—Tomographic images
  • G06T2207/10104—Positron emission tomography [PET]
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T2207/00—Indexing scheme for image analysis or image enhancement
  • G06T2207/10—Image acquisition modality
  • G06T2207/10072—Tomographic images
  • G06T2207/10108—Single photon emission computed tomography [SPECT]
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T2207/00—Indexing scheme for image analysis or image enhancement
  • G06T2207/30—Subject of image; Context of image processing
  • G06T2207/30004—Biomedical image processing
  • G06T2207/30016—Brain

Definitions

• the invention relates to a method and an automated system for assisting the prognosis of progression and assistance in the diagnosis of Alzheimer's disease for patients with mild memory impairment, or MCI (for "MiId Cognitive Impairment").
• It also relates to a method of learning such a system relating to the identification of discriminating regions in the brain and the use of these regions to refine the assistance process from new known cases.
• the invention can also be applied to assistance with diagnosis or prognosis for other neurological diseases or conditions.
• MCI MiliId Cognitive Impairment
• patients with MCI may remain stable (i.e., have little or no evolution of their memory disorders over time) or progress to dementia, particularly Alzheimer's disease.
• Imaging tools include in particular Tomography
• PET Positon or PET
• PET scan for “Positron Emission Tomography” in English
• TEM P Monophotonic Emission Tomography
• SPECT Single-Photon Emission Computed Tomography
• TEMP imaging measures cerebral blood flow, or perfusion, in the brain (123-IAMP, 99mTc-HMPAO or 99mTc-ECD), while imaging PET measures glucose metabolism (18F-FDG).
• hypoperfusion or hypometabolism in certain brain regions may be significant for a probability of conversion to Alzheimer's disease within one to three years.
• Hirao et al. (Hirao, Ohnishi et al., 2005) carried out analyzes of gouges using the logistic S PM99 in order to determine the most discriminating anatomical regimes in a population composed of 24 stable subjects and 52 converts. He thus showed a decrease of the activity within the left angular gyrus, the lower parietal cortex and the precuneus for the converted MCI. Logistic regression was then used to determine the diagnostic value of the extracted regions, with an accuracy of 73% for imaging. This accuracy was then compared with that of neuropsychological tests, which was between 70 and 78% for neuropsychological tests.
• Huang et al. (Huang, Wahlund et al., 2003) used a logistic regression using a similar approach. He also performed a group analysis using the SPM99 software, between 54 stable MCIs and 23 converted MCIs. She highlighted a decrease in activity in the left and right parietal cortex. Logistic regression was then used to evaluate the discriminating power of imaging and neuropsychological testing after anatomically segmenting each examination into 46 volumes of interest using the BRASS software. It thus obtains an area under the ROC curve of 0.75 for imaging and between 0.75 and 0.77 for the different neuropsychological tests. Finally, she combined imaging with neuropsychological tests. The performances were thus improved (area under the ROC curve between 0.82 and 0.84.
• Principal Component Analysis is a projection method that modifies the representation space, which entails a loss of information and can be a source of bias or inaccuracy ("From a set n of objects in a space of descriptors, its goal is to find a representation in a reduced space of k dimensions (k ⁇ p) which preserves "the best summary", in the sense of the maximum of the projected variance.
• An object of the invention is to assist a practitioner in his diagnosis and / or his prognosis aiming to identify a prodromal phase of Alzheimer's disease or to predict the evolution of a patient suffering from MCI type.
• the invention proposes an automated tool capable of presenting a statistical classification of examination data of such an unknown patient, or of placing it among examination data originating from a reference population of subjects. or patients of known evolutions, as described in the description and in the figures below. According to this aspect, the invention thus provides an automated method of processing spatial data representing a physiological characteristic within the brain of a patient for assisting in the prediction of Alzheimer's disease.
• the invention seeks to facilitate and make reliable the use of information concerning such a patient of unknown evolution, so as to to make the work of a person making a diagnosis or making a decision based on the real nature of the disease or disorders affecting the patient in question more reliable and easier.
• the invention seeks in particular to improve the performance of the statistical classification obtained, and therefore potentially the final diagnosis and the prognosis of evolution, in particular in sensitivity and specificity, and if possible exceeding the value of 80%.
• the tool according to the invention thus proposes to calculate and present a numerical probability base that can be used to evaluate the results of a patient studied, and / or to compare intuitively the results of the patient studied with those of a population of known and validated reference.
• the invention also proposes to improve the statistical classification obtained, in particular by a method as described below. According to this aspect, the invention further proposes to use at least one discriminant region defined according to spatial coordinates set out below.
• the invention proposes a combination of several types of examination to improve the statistical classification obtained, as set out below.
• the improvements achieved are particularly in the area of predictive reliability, ergonomics and the relevance of the practitioner's use of his knowledge and experience. In particular, it allows him to put the new case in perspective with respect to known situations, in a visual and intuitive way and in relation to a knowledge base updated and / or refined as and when progress and evolution of the cases studied.
• Another aim is to propose a method for validating and refining the prediction and positioning capabilities of such a tool, to enable the validation and / or the subsequent improvement of a method or system implementing this automation.
• a goal is also to maintain this evolution as close as possible to the real data by minimizing the biases or alterations that can be caused or amplified by the modeling, for example by modeling that is too simplistic or distorting.
• the invention thus proposes to change the proposed tool as time and use, for example from the evolution of known cases or an increase in the population of reference.
• the invention adds at least one new patient to the reference population in one or more iterations, and comprises for it an input of imaging data for that patient, as well as a recalculation of the patient. discriminant index from the new reference population.
• the invention also provides a method for automating the identification of scribing regions usable in imaging to produce the examination data and to refine the use of these regions, as described hereinafter. It may thus be easier and more ergonomic to prepare the imaging data for the calculation of the discriminant data, for example during a first implementation of the tool, or for example during a readjustment or a recalibration based on the tool using a new protocol, or a new imaging resolution, or even a new imaging technology measuring a near feature or a similar representativeness, such as a TEMP imaging pass to an imagery PET.
• the invention is based on scientific and statistical research of inventors published elsewhere, aiming in particular to differentiate automatically within the MCI population, individuals who will not present changes within 3 years, individuals who will progress to Alzheimer's disease, in particular by relying on the analysis of TEMP (Monophotonic Emission Tomography) scintigraphic images and neuropsychological tests.
• TEMP Monitoring Photonic Emission Tomography
• Such a classification can then be used, for example by an experienced medical practitioner, as an additional element of decision to decide a diagnosis from his experience of this tool.
• Automated classifications can of course be made from such a statistical classification, for example by deciding on a zone or scale limit positioned on a graphical representation of this statistical classification.
• the system can then automatically provide a classification of the patient under study, in a given category of the patient or the patient, using a graphical representation.
• the deductive medical phase comprising the attribution of the results to a clinical picture then corresponds to the setting up of such a category limit and to the choice its position relative to the reference population.
• the invention proposes a preferred embodiment, resulting from the research work of the inventors, which will be detailed later.
• a neuropsychological index (Grobert & Buschke Libre Reminder test)
• the software learns to classify the subjects using a database of MCI patients of known evolution, ie declared stable or converted by a neuropsychologist from the review of the follow-up of clinical and neuropsychological data.
• the software provides assistance to detect for new MCI subjects, so-called “unknown” or “unknown” patients, those who present a significant risk of conversion to Alzheimer's disease.
• This com m a m ent is processing im ag es with automatic extraction of regions of interest, and using neuropsychological test results, in a context of methods automatized by learning, may possibly apply to other neurological pathologies.
• FIG. 1 is a process for constituting and teaching a prognostic assistance software according to the invention in an embodiment with imaging alone or in combination with neuropsychological tests;
• FIG. 2 is a diagram illustrating the constitution and use of a prognostic assistance software according to the invention in an embodiment with imaging alone;
• FIG. 3 is a diagram illustrating the constitution and use of a prognostic assistance software according to the invention in an embodiment with imaging alone and neuropsychological tests;
• FIGURE 4 is a representation in three TEMP image views, showing a discriminating region close to the right hippocampus, obtained during the constitution of a prognostic assistance software according to the invention;
• FIGURE 5 is a representation similar to FIGURE 4, for an extracted region close to the right parietal cortex;
• FIGURE 6a and FIGURE 6b are taken from a tomographic series of TEMP images referenced H 1 to H 1 2, to which are shown in scale, in a complete H5 view (FIGURE 6a) and taken again in detail for all the views (FIGURE 6b): o on the one hand the extracted discriminant region (401) closest to the right hippocampus, obtained during the constitution of a prognostic assistance software according to the invention on the other hand, the anatomical region (602) defined as the right hippocampus; FIGS.
• FIG. 7a and 7b are taken from a tomographic series of TEMP images referenced P1 to P21, on which are represented in scale, in a complete H5 view (FIGURE 7a) and taken in detail for all the views (FIG. FIGURE 7b): o on the one hand the discriminant region used (501) close to the right parietal cortex, o on the other hand three close anatomical regions defined as being:
• FIGURE 8 is a two-dimensional graphical representation of the positioning of a reference population of 83 patients by their TEMP imaging data combined with the "G & B Free Total Recall" test, including: o patient positioning, o the scale of colors representing a statistical classification obtained by linear discriminant analysis, and o a decision boundary represented by a separator positioned on the 50% isoprobability line;
• FIGURE 9 is a graphical representation similar to FIGURE 8, for the same population minus an individual u (represented by a triangle) considered as atypical;
• FIG. 10 is a screen shot of the diagnostic aid interface of the software implementing the invention, showing a two-dimensional graphical representation according to FIG. 9 used according to the combined TEMP imaging usage option. with the test "G & B Total Free Reminder";
• FIGURE 11 is a screenshot of the software diagnostic aid interface embodying the invention, showing a two-dimensional graphical representation according to FIGURE 9 used according to the imaging option alone.
• the research work at the origin of the present invention was carried out with a reference population composed of 83 individuals all diagnosed MCI at a given time t 0 , and followed over a period of 3 years.
• a brain scintigraphy obtained after injection of a radioactive tracer, here 99mTc-ECD, by TEMP imaging was performed on these patients, as well as a battery of 57 neuropsychological tests.
• the patients were then followed over a period of 3 years by a neurologist.
• Reference population The reference population with MCI, not meeting the clinical diagnostic criteria for dementia, was recruited according to the following criteria:
• DSM-IIIR Diagnostic and Statistical Manual of Mental Disorders, 3rd Edition
• Patients were followed regularly every 6 months for 3 years.
• diagnosis was re-examined by an expert committee composed of 3 neoplasms, 3 neuropsychologists, 3 geriatric experts and 3 psychiatrists. They determined whether the clinical criteria for dementia were met using the DSM-IIIR criteria.
• dementia Alzheimer's
• a full battery of neuropsychological tests was performed 6 months later to confirm the diagnosis.
• Imaging Examination Data Examination data from imaging for different patients, either reference or unknown, must undergo standardization treatments to be comparable to each other and exploitable. These standardization treatments must be similar for the different patients, that is either identical or include patches intended to compensate for known or observed variations, for example according to the equipment used or the circumstances of data collection.
• the images are spatially re-calibrated so that they are in the same frame of reference (based on referential of Talairach) and therefore, comparable.
• the spatial recalibration was carried out using SPM2 software (Statistical Parametric Mapping) [6, 7]. It consists in applying deformations to the volume so that the anatomical regions of the brain to be recalibrated are located in the same place as those of a reference image, called "template” (average image made from 75 healthy subjects).
• SPM 2 automatically normalizes the images.
• SPM2 uses the "IRM template” (model used in Magnetic Resonance Image) to detect voxels (volume pixels) inside the cortex by thresholding values (especially with a threshold of 0.8 ). Then he scales the scintigraphy values so that the overall brain activity (represented by the infusion) is 50 ml / min. Age, gender, and imaging center were also reported as variables that could interfere with the analysis. Determination of discriminating regions
• FIG. 1 illustrates an important characteristic of the invention, which consists in selecting for the statistical treatment of the so-called discriminant brain regions, which are not necessarily confused with anatomically defined regions in the state of the art.
• discriminant brain regions are defined and named by scientists specializing in the structure of the brain, and have so far served as a minimum unit of analysis for previous work, for example in prior art documents. cited above.
• the invention thus proposes to use regions defined directly by their spatial position, of a volume if possible smaller and corresponding for example to one or more unit minimum volumes (typically voxels) as they can be distinguished by the imaging devices used.
• These discriminant regions are thus determined and identified according to the (known) evolutions of the patients of the reference population.
• discriminant regions thus forms a "mask” (102, FIGURE 1) in three dimensions which is used (123, 223 FIGURE 2 and 323 FIGURE 3) by the system according to the invention to select, or "extract” , the only imaging data to be considered as discriminant: ie the imaging data measured in these only discriminant regions, or "extracted” regions.
• This discriminant region is identified (1231 FIGURE 1) by statistical processing of the imagery data (103) of a reference population (101), and is defined by its spatial coordinates in a determined spatial reference (typically: Talairach) common at different individuals and providing a spatial coordinate system within the brain volume.
• This determination method comprises the following steps:
• 103 representing a spatial distribution within the brain of at least one quantitative cerebral physiological characteristic (typically: perfusion or metabolism), observed by three-dimensional imaging (typically: TEMP or PET) according to a protocol common to said reference patients;
• at least one quantitative cerebral physiological characteristic typically: perfusion or metabolism
• three-dimensional imaging typically: TEMP or PET
• a spatial normalization 121 by registration or deformation of the images to provide a spatial representation of the brain that conforms to a determined spatial reference system
• said reference providing a spatial coordinate system within the cerebral volume, which also has a non-parametric ratio adjusting the set of values of the physiological characteristic noted. within the brain of this patient, so as to provide an overall quantitative value of said physiological characteristic that is consistent with a specific quantitative reference, common to the different patients (here: 50ml / min); group comparison 123 (here with SPM2 software) values of the physiological characteristic measured between at least two groups of reference patients that have evolved differently, for each spatial zone or voxel observed, thus providing at least one so-called discriminant region 401, 501 defined by spatial location according to this spatial reference.
• FIGURE 4 and FIGURE 5 show the two discriminant regions 401 and 501, respectively extracted from the imaging data, in the right hemisphere and near the anatomical regions of the hippocampus (FIGURE 4) and the parietal cortex respectively (FIGURE 5 ). These regions are consistent with the topography of known lesions in Alzheimer's disease.
• FIG. 6 and FIG. 7 thus illustrate, for these same two regions 401 and 501 discriminant "extracted” according to the invention, the localization of the closest anatomical regions as defined according to the AAL ("Automated Anatomical Labeling") standard and defined manually on the SPM2 IRM template (cf.Tzourio-Mazoyer N, Landeau
• the invention thus proposes to use at least one discriminant region having spatial coordinates (according to the atlas proposed by the Montreal Neurological Institute ", close to that of Talairach) including at least one of the following sets of coordinates (the lines in bold represent the most significant part of the region):
• Puncorrected indicates the level of significance of the test; PFDR- ⁇ - is a corrected Puncorrected value that takes into account the number of tests performed; T is the statistical value of the test; x, y and z are used to localize the regions in the repository used.
• the invention proposes to use one or more discriminant regions (401, 501) that do not correspond to any of the so-called anatomical regions defined in the state of the art.
• the invention proposes to use at least one discriminating region comprising at least two zones located in different anatomical regions and comprising less than 75% of each of said different anatomical regions.
• such a discriminating region 501 comprises at least three zones situated in the following anatomical regions:
• FIG. 6 and FIG. 7 thus illustrate two discriminant regions according to the invention, here called “hippocampal discriminating region” 401 and “discriminatory region of right parietal cortex” 501, as identified and used by the invention in the embodiment described here (TEMP imagery). These figures illustrate the location of the anatomical regions closest to the discriminant regions extracted according to the invention. These anatomical regions are respectively the region 602 of the hippocampus according to its “anatomical” definition, and respectively the “anatomical” regions that are close to the "angular gyrus” region 702, the “lower parietal cortex” region 703 and the "gyrus” region. supra marginal "704.
• the two discriminating regions 401 and 501 in the right hippocampus and in the right parietal cortex were extracted following the analysis of groups 123 carried out in SPM2, and retained for the constitution of the mask 106.
• included in this mask 106 are the regions of the hemisphere which are symmetrical with discriminating regions 401 and 501 located in the right hemisphere.
• the different combinations of these variables can be used in the process according to the invention as well as the different combinations obtained by grouping certain variables (for example, by calculating the average activity over the whole of the left and right hippocampus).
• the combination chosen is a single variable obtained by grouping, calculating the average of the activity on the four extracted regions, ie the regions extracted in the hippocampus and the parietal cortex, at right and left.
• imaging makes it possible to obtain a figure that reflects the average activity in this region, through the measured radioactivity.
• This figure is proportional to the blood flow (infusion), but does not provide an absolute value for this blood flow. It can be expressed as a percentage of the average recorded for the same patient: for example 70% or 120% of the overall mean cerebral value of the patient.
• This figure can also be reduced to a physiological value, as in the case of the SPM2 software which assigns the standard value of 50 ml / min to the average value (100%) of the activity recorded for each patient, for example in the quantitative standardization phase (122, 221, 321).
• Classification As a function of their subsequent evolution 104, the reference patients are classified by a discriminant analysis statistical treatment, providing a numerical probability scale. Discriminant analysis will calculate the probability that a given individual belongs to each of the classes present. To calculate this probability, one uses the information of the learning base (for example, the average or the variance).
• this classification provides a first numerical scale 108 of conversion probabilities.
• This scale 108 is said to be “physiological” and is based on the "physiological" isocrine inhibitory indices 105 of the reference patients, that is to say obtained solely from their imaging data 103.
• Such a scale can to be graphically represented 124 in the form of a one-dimensional scale, for example as the scale of color gradients shown vertically to the right of FIGURE 8, FIGURE 9 and FIGURE 10 and graduated from 0 to 1.
• the invention thus proposes a computer tool implementing an automated method for developing a probability scale 108 (FIG. 1) for conversion to Alzheimer's Disease for patients with mild cognitive impairment, or MCI (for "MiId Cognitive Impairment”).
• This process comprises the following steps:
• At least one discriminant region 401 and 501 defined by its coordinates in a determined spatial reference system (typically: Talairach), common to different patients 101, 201, 301 and providing a spatial coordinate system within the brain volume ;
• a determined spatial reference system typically: Talairach
• extracted imaging data representing said discriminant region, within digital data 103 representing a spatial distribution of at least one characteristic quantitative cerebral physiology (typically: perfusion or metabolism), observed by three-dimensional imaging (typically: TEMP or PET) according to a protocol common to said patients;
• characteristic quantitative cerebral physiology typically: perfusion or metabolism
• three-dimensional imaging typically: TEMP or PET
• this process also comprises, alternatively or successively:
• a spatial standardization step 121 of the imaging data 103 arranged so as to provide for all these patients 101 a spatial representation of the brain that conforms to a determined stable spatial reference from one patient to another, said reference system providing a system spatial coordinates within the brain volume;
• a step of quantitative normalization 122 of the imaging data including an adjustment of the set of values of the physiological characteristic found within the brain of each patient 101, arranged so as to provide an overall quantitative value of said physiological characteristic which is in conformity with to a specific quantitative reference, common to the different patients (here: 50ml / min for infusion).
• the method may then further comprise calculating the discriminant indices 105 of a plurality 100 of reference patients from their measured values for the physiological characteristic, and the graphical positioning 124 of said reference patients on a one-dimensional probability scale 108. converting from said discriminant index.
• the inventors' research also focused on a combination of imaging data with the results of one or more neuropsychological tests. This combination has shown an improvement in the performances obtained and may or may not be included in the tool proposed by the invention.
• Two types of test were selected for their performance, among 57 types of neuropsychological tests that were performed for each patient 101 of the reference population 100. In the same way as for imaging, a study of the discriminating power of each test was performed (by monovariable classification) to select these two tests.
• a test is chosen from the two Grober and Buschke tests, which are the "free call” type test or the "cupped callback” type test.
• results 102 of this test are associated with the physiological discriminant index 105, to be used as coordinates 126 to position that patient on a two-dimensional graphical representation. As illustrated in FIGURE 8, the combination of test results
• the imagery-only one-dimensional conversion probability scale 108 then provides a numerical scale 109 of conversion probability that can be described as "two-dimensional” or "composite". That is, the numerical values of this probability scale 109 are positioned within a two-dimensional space, here by the color variation within a rectangular array 800 along two perpendicular axes 812 and 813 representing the test data 102 and respectively the discriminant index 105 obtained for the imaging data 103. Representing on this scale the bid i mension nel the 109 associations 125 of data 102 and 103 corresponding to all patients of reference 101, a two-dimensional graphical visualization 110 of the reference population 100 is then obtained with respect to the bidimensional probability scale.
• the reference patients are shown in FIGURE 8 by the points 807, 808 positioned in Table 800.
• the 807 square dots represent those of reference patients 101 who converted to Alzheimer's disease, and the 808 round dots represent those which remained stable.
• the discriminant analysis step also relates to a numerical result 103 of at least one same neuropsychological or cognitive test performed by each of the patients. 101 reference.
• the invention also includes the generation of a graphical representation of a two-dimensional distribution probability scale 109 based on a part of the value of the discriminant index 105 for each variable. the physiological characteristic and on the other hand the value of the cognitive test result 103.
• the invention proposes the calculation of the discriminant indices of a plurality of 100 reference patients for the imaging data 103 and the graphical positioning 126 of these reference patients on a distributed conversion probability scale 109. on two dimensions depending on a part of the value of the discriminant index 105 for the physiological characteristic and on the other hand of the value of the cognitive test result 103.
• the invention When learning such a computer tool, the invention also proposes a step of calculating at least one statistical performance indicator using the "leave-one-out" method.
• This validation method consists of successively extracting each individual from the database. The model obtained by learning on all but one data is then tested on the extracted data.
• the invention thus proposes a combination of imaging and neuropsychological tests, which should therefore make it possible to improve the results of each modality.
• the preferred embodiment of the invention thus combines the best performing variable (average of the activity of the four extracted regions) for imaging with the most effective neuropsychological test (G & B Free Recall).
• the following table represents the confusion matrix obtained through linear discriminant analysis, by combining the most efficient variables for imaging (mean of the activity of the four regions extracted) and the neuropsychological tests (G & B Free Recall), after validation in leave-one-out:
• FIG. 9 illustrates a "two-dimensional" numerical scale of conversion probability 109b of the same type as that of FIG. 8, but without taking into account a "converted" individual 809, which is particularly atypical within the converted population (red squares ) for the following reasons.
• This atypical individual has a high activity in the regions examined by imaging, an activity that is close to stable individuals and not to other converted individuals.
• the G & B Free Recall results are consistent with those of other converts.
• his clinical evaluation by a neurologist showed that his evolution was compatible with Alzheimer's disease.
• this individual is the only convert present in an area of the numerical scale where the stable individuals (green circles) are mostly represented.
• the following table represents the confusion matrix obtained by linear discriminant analysis, by combining the most effective for imaging (average of the activity of the four extracted regimens) and neuropsychological tests (G & B Free Recall), after leave-one-out validation:
• the statistical rankings made on this reference population are used as the engine and database of a software executed by a computer system, and implementing a method of assisting prediction for patients whose evolution is not still known.
• a computer system In the preferred mode of realization, several options are available in the computer system and implement several alternatives or variants for this prediction assist method.
• one or the other of the options may be chosen.
• the system according to the invention then comprises means arranged to classify the examination data of a patient optionally:
• the combination of the test results 103 with the imagery-only one-dimensional conversion probability scale 108 then provides a two-dimensional numerical scale 109 of conversion probability.
• FIG. 2 illustrates the method implemented by the prediction assistance option 229 with imaging data only.
• This positioning 224 can be done numerically, simply by obtaining an index 205 on a purely digital scale. It can also be done visually, by graphically positioning 224 an indicator for this patient to be predicted on the same graphical scale as patients 101 of the reference population 100.
• the invention provides an automated method of treating imaging data 203 representing at least one cerebral physiological characteristic in a patient 201 with mild memory impairment, or MCI (for "MiId Cognitive Impairment"), in view of assistance in predicting the onset of Alzheimer 's disease, or conversion. This includes the following steps:
• - Normalization processing 221 of the image data obtained comprising: o on the one hand a spatial normalization of the images obtained, so as to provide a spatial representation of the brain according to a determined stable spatial reference of a patient to the other, said frame providing a spatial coordinate system within the cerebral volume; and o on the other hand a quantitative normalization adjusting the set of values of the physiological characteristic found in the brain of said patient, so as to provide an overall quantitative value of said physiological characteristic that is in accordance with a specific quantitative reference, common to the different patients (here: 50ml / min);
• a classification method with supervised learning preferably a linear discriminant analysis 223, of the functional characteristic values recorded in a selection of one or more voxels forming at least one predetermined discriminant region 401, 501, 106 defined by its coordinates within said spatial reference system, said discriminant analysis providing for said patient a value 205 of said physiological characteristic that can be compared with a plurality of reference values identified and calculated for patients 101 of known evolution reference.
• FIGURE 3 illustrates the method implemented by the 329-based prediction assist option based on test data including 303 imaging data and 302 results of a neuropsychological cognitive test for 301 patient. evolution 304 to predict.
• the patient to be predicted can then be positioned 326 on a two-dimensional graphical representation comprising a two-dimensional numerical scale 109 of conversion probability, on which reference population 100 can also be included.
• FIG. 10 illustrates an interface screen of the software implementing the invention.
• This screen includes a computer window 9 displaying the same scale as the one shown in FIG. 8 for the constitution and the graphical representation illustrated in FIGURE 8. learning the help tool.
• This screen also includes a frame 902 of choice of the prediction option, allowing to select or not the use of a type test
• This frame also comprises an input field receiving the result 302 of the test chosen for the patient to be predicted 301.
• the software displays in the graphic frame 900 a point 901 (here star-shaped) representing the patient to be predicted 301 and positioned along a vertical axis 912 to the test result 302 and a horizontal axis 913 for the discriminant index 305 from the imaging data 303.
• a point 901 here star-shaped representing the patient to be predicted 301 and positioned along a vertical axis 912 to the test result 302 and a horizontal axis 913 for the discriminant index 305 from the imaging data 303.
• the bidimensional probability scale 109 as defined previously for the reference population 100 makes it possible to easily evaluate (here by the value of the color gradient at the level of the 901) a conversion probability value for this new patient 301. This value is also displayed numerically by the software in a display field 905.
• this patient 301 to be predicted thus quickly and intuitively provides a practitioner or a user with a basis for deciding on a prognosis or diagnosis, for example by using his experience or according to a diagnosis. clinical strategy.
• all or some of the reference patients 101 are also positioned in the graphical representation frame 900 according to their own results.
• the converted patients are here represented by a 907 square, and the stable patients by a 908 round.
• This graphical distribution of the reference population 100 thus makes it possible to easily and intuitively display the position 901 of the new patient 301 among the positions 907, 908 of the reference patients.
• This visualization thus makes it possible not only to quickly locate the distances between them, but also to evaluate simply whether the patient to be predicted 301 is in a graphic zone of the frame 900 where the reference patients 101 are numerous or not, which can give an intuitive insight into the specific reliability of this 901 prediction.
• the operator can intuitively visualize that the prediction provided may not be as reliable as the overall performances of the represented model.
• the representation of the scale here the color gradient
• the visualization of the reference population can be performed concurrently or separately, depending on the options or the embodiments.
• this screen can also be used to calculate or display a prediction on imaging alone, for example according to the choice checked by the user in the 902 selection frame of the test.
• predicted patients 201 and reference 101 are displayed in a linear fashion.
• the reference patients are furthermore divided into several lines to indicate in a visible way whether they are patients having converted 907 or not 908, 909.
• the subjects can also be displayed on the vertical scale. 108 on the right, for example in the form of a vertical slider for the predicted patient 201 and a shaded more or less dark depending on the concentration of reference patients 101 at each height of this scale.
• the graphic frame 900 can also simply be modified to have vertical uniformity and to represent the one-dimensional scale 108 on the horizontal axis 913.
• the method of assisting the prediction according to the invention further comprises the following steps: - association 327 of the value 305 of physiological characteristic, obtained for the evaluated patient 301, with at least a quantitative value 302 cognitive performance from at least one neuropsychological test performed by said patient;
• the method of constitution can then comprise one or more iterations of adding at least one new patient of known evolution to the reference population 100, comprising on the one hand an input of imaging data for said patient and secondly a recalculation for the new reference population of the numerical function of obtaining the discriminant index for the conversion from the value of the measured physiological characteristic.
• This learning can for example be realized as and when by the operators of the assistance system:

Landscapes

• Engineering & Computer Science (AREA)
• Quality & Reliability (AREA)
• General Health & Medical Sciences (AREA)
• Medical Informatics (AREA)
• Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
• Radiology & Medical Imaging (AREA)
• Health & Medical Sciences (AREA)
• Computer Vision & Pattern Recognition (AREA)
• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Theoretical Computer Science (AREA)
• Nuclear Medicine (AREA)
• Measuring And Recording Apparatus For Diagnosis (AREA)
• Magnetic Resonance Imaging Apparatus (AREA)

Applications Claiming Priority (2)

Publications (1)

Family

ID=40289214

Family Applications (1)

Country Status (6)

Cited By (1)

Families Citing this family (25)

Family Cites Families (7)

• 2008
  • 2008-05-15 FR FR0853158A patent/FR2931281B1/fr not_active Expired - Fee Related
• 2009
  • 2009-05-14 JP JP2011508987A patent/JP2011521220A/ja active Pending
  • 2009-05-14 WO PCT/FR2009/050905 patent/WO2009150349A2/fr not_active Ceased
  • 2009-05-14 EP EP09761900A patent/EP2277143A2/fr not_active Withdrawn
  • 2009-05-14 US US12/990,561 patent/US20110046451A1/en not_active Abandoned
  • 2009-05-14 CA CA2721854A patent/CA2721854A1/fr not_active Abandoned
• 2014
  • 2014-03-05 JP JP2014043048A patent/JP2014145770A/ja not_active Withdrawn

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events