JPS6088539A - automatic diagnostic equipment - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an automatic diagnostic device that performs logical diagnosis processing 4f regarding metabolic abnormalities based on analysis data from an analyzer and specimen information +11, and extracts names of metabolic abnormalities.

In recent years, there has been remarkable progress in research into inborn errors of metabolism. The congenital metabolic disorder is caused by a genetic malfunction due to the deficiency or inhibition of enzymes that act as catalysts at each reaction step of the metabolic pathway in the body. This metabolic abnormality is highly effective if treated early, so the significance of early detection is 441st. It is estimated that there are more than 2,000 metabolic disorders, and it is believed that several hundred of these can be diagnosed using various analytical methods such as J, amino acids, organic acids, sugars, and slaroids. (I'm being admired.

The present invention provides a novel device that automatically diagnoses metabolic abnormalities based on analysis data and sample information from existing analyzers.

The present invention is an analyzer, in which a healthy value table showing the healthy range (normal range) for each item is stored. A second memory Cornit stores an item/disease name table that indicates the degree of involvement with the item and the rJ weight, and abnormality model weights based on the items related to each disease name and the degree of involvement of the item with the disease name. The third memory unit 1-1 in which the disease name/term [1 table is stored shows the model symptoms that are related to each disease name and the weight that indicates the degree of relationship of the model symptoms to the disease name. A fourth memory in which the disease name/symptom table is stored, and the normal value table is referenced based on the normalized sample analysis data from the analyzer and the sample information to determine the abnormality of the C sample. Detect items and calculate their degree of abnormality. For the detected abnormal items, refer to the item/disease name table to extract the name of the suspected disease of the specimen and calculate the degree of suspicion of the name of the suspected disease. Regarding the name of the suspected disease, refer to the disease name/item table above to find the abnormality 11'i ITI that matches the sample and the model listed in the table. The abnormality degree of the abnormality of the abnormality of the abnormality item and the weight of the abnormality model of all the abnormality items of the corresponding del. , the abnormality degree of the abnormality Ifi which matches the sample and the model listed in the table, the weight of the logical model of the abnormal term 11, and the sum of the weights of the logical model of all abnormal items of the model. Based on the logical model matching degree calculated based on the results, the stem is identified, and the name of each suspected disease extracted is then referred to the disease name/symptom table.
An operation that calculates the degree of symptom model matching based on the degree of the symptom that matches the model listed in the pull, the weight of the model symptom for that symptom, and the sum of the weights of all model symptoms in the model. This is an automatic diagnostic device consisting of a .

FIG. 1 is a schematic diagram of an automatic diagnostic device shown in C as an embodiment of the present invention. The automatic diagnosis method of the present invention will now be explained in detail with reference to the figure. In this example, an amino lI tablet analyzer was used as the analyzer.

In the figure, 1 is an amino acid analyzer (・Sample analysis data from this analyzer is stored in a central illumination system (hereinafter referred to as CPU).
According to the command, it is stored in Dismono Shichiri 3. The CP
U reads the sample analysis data from the disk memory, takes the data, that is, the data of each peak intensity of the chromatogram of the amino acid analyzer, and normalizes the data of the retention time. That is, depending on the elution conditions, the retention time of the same eluted components may be the same depending on the elution conditions. I\'tZ :L Sometimes (,
-r, each? Create a Δoriginal chromatogram with the same C retention time as the Ff output component, and determine which peak retention time of each peak in the 117-togram is determined by which of the peaks in the A-retention chromatogram. Normalize the peak retention ratio based on whether the time is within the acceptable range.

Then, according to the instructions of the CPU, the standardized minutes 1
1 data is input to the storage device 4 as a data file. The data 7 also includes manual information such as age, age, patient information (pre-existing conditions, symptoms, medication regimen), and sample size 8 (specimen N). (1,)′!
Sample information of i is input.

Next, CPtJ2 reads the data file 4 /Jl l
The standardized analytical data and sample information are extracted from the bladder and the eluted components stored in the first memory unit [5]
Item name -) For healthy 1 direct box 1 of fij (j[;'i
t Refer to the healthy value table showing the normal values for each age group, and based on the standardized analysis data (detect abnormal items and measure their differences, i.e., CP LJ 2
C converts the peak intensity of each item of the standardized analysis data into a histogram, and simultaneously inputs the healthy value range of each item read from the '1st storage unit 5 into the histogram.
do. Also, the average value of the population of healthy data is included in the histogram. Then-1, the His I-gram is displayed on the display device 6 (cathode ray color and/or J printer) as shown in FIG.

At d3 in Figure 2, the number on the horizontal q1+ (to o, > (1)
.

(2), (3), ...... are peak numbers,
Among them, (1) is thread A nin (-l-1+r), (2>
is j7 lanine (△la), (3) GJ glycine (G
lyn, (l is valine (Vale), (5) is mebA nin (Mat).

This is the item... The mark on the vertical axis indicates the peak intensity, and the area between the markers (-) and 1 is the peak intensity.
1: value range, and the x mark indicates the average value of the population of healthy data, respectively.10 From the figure, the peak intensity is not within the normal range. Next, as shown in Figure 3, the standard deviation of the peak intensity of each item from the average value of the population (marked on the vertical axis) and the standard deviation of the normal value of each Jr. 111 line) is displayed. In addition, the solid horizontal axis is the population's ゛(j mean (x), the dashed horizontal line is 2a, -2a <σ2@I'; 1 variance)C
be. Next, if the peak intensity of each term 1] of the sample is larger than the average value of the n1 population, what is the ratio between that (1r1 and R7Tf value)? This ratio is determined in advance based on the metabolic efficiency in the body (according to the Qrs constant threshold set for each item (-5
, -4, -3, -2, -1° It is leveled into stages of 0.1, 2, 3, and 4.5.

This leveled bean paste is displayed as a histogram in Fig. 4 as an abnormality level. This platform, I(t;+:';
41+ The difference 7i of the term L1 (1 wealth item) that is within the range is o. Furthermore, CP tJ 2 extracts the abnormal items, and as shown in C1 Figure 5, they are sorted in descending order of abnormality degree and -
are displayed on the display device 6 in descending order of abnormality level. .

The above processing is called the primary diagnosis1.Next, the CPU 2 performs the C1 secondary diagnosis on the iP items detected in the primary diagnosis.
Extract the name of the suspected disease by referring to the IC item/disease name table stored in the memories 1 to 7 and write out the first page of the initial suspect in line 4r.
cormorant. As shown in Figure 6, each item (threonine (1-br>, valine (Vat),
Names of diseases related to alanine (Ala>, ・・・・・・・・・) (12, 7, 75, /l, ・・・・・・・・・
・3.6.12.3. . . . coded as .
3 ,・・・・・・Δ21゜△22.・・・・・・
...) is memorized and -( exists.

CI) U2 refers to the daple and determines the 3nth abnormality detected in the first diagnosis (Thr, Val, Δla,
△ “S.

Extract all the disease names that are related to each (MeL, l)n), and as shown in Figure 7, among them, the closely related ones are divided into 3
A command is sent to the display device 6 to select one item and display it on the lead of each purchased item. At this time, as shown in FIG. 7, the specimen information of abnormality level +4 (+) Gal and abnormality level -5 of lactic acid are inputted. In addition, the real name of the disease, which has been converted into =1-t, is written on a part of the display device so as to cover it. Next, for each suspected disease name extracted in this way, initial suspicion 1 is used as a line.

i is different t'A term day, 1 is disease name, / is different degree, Φ is,
If we take AmaX as the most popular among the above data and Aij as the disease name related to each item (specific move I), then the suspicion level of disease name J is 1 Dl, and i is 1) D L , i-Σ(AijX l 7i l)
/ (AllaX x (specimen abnormality) Ii tEl number)
]・・・・・・・・・(1) For example, disease name 12 (strength de)
The degree of suspicion is DD112-△+ + X5+Az 2 x
4-1-△3, ×3+・・・・・・・・・/(Δmax
xε3).

In this way, disease name 12.7, 75.・・・・・・・・・
The CPU 2 displays the names of the suspects on the display device 6 in the descending order of the suspects as shown in Figure 33. (3) In this case, the names of the illnesses are displayed using their real names. A'1 is called the secondary diagnosis.

Next, CI) LJ 2 is extracted from the second consultation rDi ``C 1.: Name of each initial suspected disease (maple, phenyl...
(...), by referring to the disease name/slH table stored in the third memory Jnira 1-8,
Sequentially, the extracted disease names (maple, phenyl, etc.)
...), and arrange them into the normal items/suspected disease name list 1 for the specimen as shown in Figure 7 above.
An abnormal item model list of disease names is displayed on the display device 6 (the list shown in FIG. 9 is a list of )j]-de). Let's take the disease name Kaede as an example and measure the degree of suspicion related to the abnormality item, that is, the 15 suspicions.If the degree of suspicion is measured, then the abnormality degree model - several animals logic'E Measure the del agreement.Anomalous degree model - Some animals look at the degree of agreement between the specimen and the actual temperature model, so logical model - Some animals measure the degree of agreement between the specimen and the theoretical model. It is something to look at.

In the disease name/item table, as shown in FIG. 10, for each disease name (maple, phenyl, ...), the related I page II (Thr, Val, ...・・・
..., Mot, ......) and the degree of relationship of each item to the disease name 1J! i′I: Always 1 (l model'J only (B+ + , B+ 2, −−, −, r3+ n
.

・・・・・・・・・・・・＋ B21.・・・・・・・・・
) and the logical model car (C1l+CI2+"""+C
1n+"""+C2+...) are respectively stored.

The abnormality model agreement degree AG1j for the disease name j is 1 night when Sl is the sample and Del is the same.
The sum of the abnormality degree models of the abnormality degree models for each item that is on the model list and (Met), Mi
When the sample is sampled, the abnormal term 1, which is consistent with C, is 5 of 1.
``4X':''; The weight of I sad Del/' [-Each different Hatake Jf that is higher on the Del list! If we take the sum of the weights of the eye abnormality model, ΔG1j = 1− (Σl Si −1v
li l ) /2 (2). For example, the disease name Kaede is 1 (lri/
(13t I 10B+ 2 10B+ +t) -B++
/(I3+++13+2+r3+n)1114/(B+
+ +B+ 2 +B+ n ) -Bl 2 (1
3+ + 1 [3+ 2 -1-BI II) l-
1-I AI (13+ + -11'(, 2+B+n
) l:3+n N3++ +B+2-l-13tn)
1]/2.

Next, the logical model of disease name 1-several animals AG2j is
In i, the weight of the anomaly degree model is the weight of the logical seven dels, and in the above 1yli, the difference 7:37 fu'
If we take the weight of the model and the weight of the logical model as -[1, then ΔG2j-1-(Σl Ei −-I'i l ) /
2・・・・・・・・・(3) For example, regarding the disease name Kaede (ha1 (15/
<C+ + +C+ 2 +CI I+ > -C+
1/ (C+ + +C12 toC+ n) l +l
4/(C+ + +C+ 2 +C+ n ) C1
2/ (C+ ++C+ 2 +C+ n ) 14-
l 4/ ('(:+ + -1-(, +2→-C+
n ) C+ n / (C+ + +C+ 24-C
+n)l)/2.

Regarding the above 2 degree degree of concordance" (is the second examination N
Each disease name extracted by Ji is sequentially determined.

These - several animals take a value between 0 and 1, where O means a complete mismatch and 1 means a complete match. Different Hatake model - In the case of several animals, the closer it is to 1, the closer it is to the actual model, and the closer it is to O, the further away it can be diagnosed. In addition, in the case of a logical model of several animals, the closer it is to 1, the closer it is to the theoretical model, and the closer it is to O, the farther it is from the diagnosis 111. These two matching degrees are determined by the instructions of the CPU 2, respectively.
The disease names and disease name σ pieces of several animals are displayed in descending order on the display device F16 (see FIG. 11). The above operations are called tertiary diagnosis.

Next, CPU2 is extracted in the second examination and 1. Regarding the suspected disease name (power) - de, f1nir,... model symptoms (Δ
, ○1ro, ◎) and call the specimen's symptoms (Δ, ×, ☆, ◎) and their severity (+2.+3.-1
-2. +4) is displayed on the display device i'l6. In Figure 12, the actual name of the symptom is symbolized and displayed. In the disease name/symptom table, as shown in Figure 13, each disease name (Kaede, ] ■ Niru...
...), model symptoms related to each disease name (Δ,○
2 mouths, ......, ◎) and the degree of involvement of the disease name of the symptoms (1)] 1, DI2.1
+ 3 , - ・=, DI n , --l), D
21,...) is written in 1Q. Then, for each suspected disease name of the C1 specimen, refer to the disease name/symptom table 9 and select the number of symptoms. Disease name jσ month, i-like coincidence antiΔG
3j is J3 with 1-11 as the disease name, and symptoms i (e.g., disease name) that match in the specimen, model, etc.
(Ha△do◎〉's culm I sad/The sum of the weights of all C model symptoms related to the disease name J, Qi for the disease name j (the weight of symptom 1 that matches the sample and model) If we take the weight of the symptoms/the sum of the model symptoms at ℃ related to the disease name J, then Δ G3j=1 ~ Σ (l 1-1i −Qi
l ) /2 (4).

For example, regarding the disease name Kaede, 1 - (12, / (DI l → -1) + 2 I
In-D++/(D++ +D+
2-11)+341)+n)-+4/ (Dt+-
1-DI 2-tl)+ 3 +D+ n) l)+
n/ (DI 1+I)+ 2 +D+ 3 +1)
I n ))/. This symptom - some animals are on second visit WI
C. Each of the extracted suspected disease names is requested. The coincidence hemp takes a value from 0 to 1, and the closer it is to 1, the closer it is to the theoretical model, and the closer it is to 0, the more distant the culm can be diagnosed. As shown in FIG. 14, the degree of coincidence requested in this way is displayed on the display device (5) in order of disease name (11°) in accordance with the command from CI (1).

Next, Cl) U 2 is the name of the disease determined in the second consultation...1.
DI for initial suspicion) Lj, abnormalities found in the tertiary diagnosis, jii
.

degree model - several animals ΔG 1 'J, logic [del match hemp Δ
A comprehensive evaluation (1
rITVj and 'C's best Jru. These time constants α, β, and γ.

δ is an empirically determined parameter (L!y) 1, which can be expressed as:
-γ (AG2j)4- δ (AG3j > ...
... ... (1ra) and 4T are used for comprehensive diagnosis of disease name extraction 3. Figure 15) shows the total value of each disease name extracted as 0 in this way (1r1 is the highest value) The names of the diseases are displayed in order on the display device 6. In addition, in FIG. -Several animals C9 symptom concordance of 1 is avoided by bargrano display.

The judgment criteria of this system are all based on the numerical values in each table, and it is separated from the system's program file I-1-, and it has a table backup file I! (Therefore, the user can set any diagnostic logic by modifying the above full contents.Also, in the above example, amino acid analysis data was taken as an example, but organic acid Analytical data such as sugar data, diffusion data, and guanidine data can also be used.

With the present invention, accurate disease diagnosis is possible.
Also, the degree of matching of the seven full patterns is fil i
The degree of suspicion of the disease name of the Q-coded C1 specimen becomes clear, and it is very effective for diagnosis.

[Brief explanation of the drawing]

Fig. 1 is a schematic diagram of an automatic diagnostic device according to an embodiment of the present invention, and Figs. . 12 Amino acid analyzer 2: central control till '4 R'l (CPU)3
:Disk memory 4:Data storage unit 5:First storage unit 1'6:Display device 7:Second storage unit 8:Third storage unit 9:Fourth storage unit 1Special 5'[ Applicant L1 Hondenshi Co., Ltd. Representative: Mr. Ito Figure 3 Figure 5 Figure 7 Figure 8 Figure 9 Figure 1) Figure 12 Figure 14 Figure 19 Figure 16

Claims (2)

[Claims] (1) Analyzer, healthy 111 range for each item (normal range)
The first storage unit 1~ stores a 4iP normal value table, and the second storage unit stores an item/disease name table indicating a disease name related to each item and a weight representing the degree of involvement of the disease name with the item. A third memory stores a disease name/item table showing the items related to each disease name and the degree of relationship of the item to the disease name, the abnormality degree model weight, and the logical model weight. The unit indicates the model symptoms associated with each disease name and the degree of relationship of the model symptoms to the disease name, and sends the model symptoms to the fourth storage unit 1 in which the disease name/symptom table is stored, and from the analyzer. Based on the standardized sample analysis tailor and the sample information, the degree of abnormality of the detected abnormality in the sample is determined by referring to the above-mentioned healthy value table, and the above-mentioned item/disease name is determined for the detected abnormal item. By referring to the table, extracting the suspected disease name of the specimen and calculating the degree of suspicion of the suspected disease name, and referring to the disease name/item table for each extracted suspected disease name,
The degree of suspicion related to abnormality Jzu I [1 is determined by calculating the degree of suspicion regarding the abnormality 7j' + Ia of the abnormality item that matches the model listed in the specimen and the model listed in the daple, the weight of the abnormality degree model of the abnormality item, and the weight of the abnormality degree model of the abnormality item, and the weight of the abnormality degree model of the abnormality item, and the weight of the abnormality degree model of the abnormality item Anomaly degree model matching degree calculated based on the sum of the weights of the abnormality degree models of all abnormal items, the abnormality degree of the ¥11 normal item that matches the sample and the model listed in the table, and the logic of the abnormality item. - calculated based on the weight of the del, and the logical model matching degree calculated based on the sum of the weights of the logical model of all abnormal items of the model, and further, for each extracted suspected disease name, the disease name/symptom table is By referring to the sample and the model listed on the table, the degree of the symptoms that match and the degree of the symptoms, the symptoms based on the sum of all the model symptoms of the model. An automatic diagnosis device consisting of a performance device that determines the degree of model matching. (2) The analyzer goes to the first storage unit 1 in which a healthy value table indicating the healthy value range (normal range) for each item is stored;
The second storage unit stores the item/disease name table, which displays the disease name related to each item, the degree of relationship of that disease name to the item, and the weight. The abnormality model weight and logical model weight indicating the degree of relationship of the item to the disease name are transferred to the third storage unit 1 in which the disease name/item table is stored, and the model symptoms and their associated symptoms are respectively related to each disease name. To the fourth memory unit 1 in which a disease name/symptom table indicating the degree of involvement of the model symptoms to the disease name and the weight is stored, and the standardized sample analysis data from the 417 necks above and the sample information. Based on the above-mentioned healthy bullet value table, detect the bre:base item of the specimen and calculate its abnormality level, and for the detected 5'4 normal item, refer to the above-mentioned item/disease name table to determine the suspect of the specimen. By extracting the disease name and calculating the degree of suspicion of the suspected disease name, and referring to the disease name/item table for each extracted suspected disease name, the degree of suspicion related to the abnormal item can be calculated based on the specimen and the table. The anomaly terms that match in the model [1 Tatsumi 1:1 degree] Based on the weight of the anomaly degree model of the anomaly item, the sum of the weights of the 'A 78 degree model of all the anomaly items of the model Calculated difference: 1 soldier who matched Shima Shichidel and the specimen) -
The abnormality degree of the abnormal item that is listed in the pull and is alone in C on Del, the weight of the logical model of the abnormal item, and the sum of the moves of the logical model of all abnormal items of the model. The extent of the symptoms that match between the specimen and the model listed in the table is determined by referring to the disease name/symptom table for each extracted suspected disease name. The symptom model match is calculated based on the width of the model symptoms of the symptom, the sum of the weights of all model symptoms of the model, and the initial suspicion of the disease name calculated above, the degree of abnormality model match, etc. An automatic diagnostic device comprising a stop device that calculates a linear function that covers the logical model matching degree and the symptom model matching degree as variables as a comprehensive evaluation area.