DESCRIPTION

"Control quality system for biological samples"

The present invention relates to a control quality system for biological samples.

The present invention also relates to a control quality system for biological samples.

In particular, the present invention relates to a control quality system and method for biological samples of the type usable in the pre-analytical phase, i.e. upstream of laboratory analyzes on the biological samples themselves, for example on Dried Blood Spots (DBS) or on test tubes containing biological samples.

As is known, the use of dried blood samples absorbed on special papers, the "Dried Blood Spots" (DBS), for diagnostic purposes, has been known for decades and has been mainly focused on the diagnosis of infectious diseases, especially for neonatal screening of hereditary metabolic diseases. Only recently the use of DBS has expanded to a variety of new and innovative applications.

The entire pre-analytical phase, which includes the sampling and processing of the biological samples to be 2 analyzed, and in particular of the DBS, before the final analysis, is currently neither standardized nor automated.

Blood chemistry analyzes are currently carried out in analytical laboratories. The medical or paramedical staff take a blood sample from the patient, and the blood sample taken is stored and sent to the analysis laboratory. The laboratory performs the required analyzes and issues a report which must then be transmitted or transported to the patient and his / her doctor. The acquisition and storage of the biological sample can take place by sampling on a dry matrix or by means of DBS samples. DBS samples are suitable for analysis only when the volume of blood absorbed on the support, the measurement and the positioning of the dry sample with respect to the support are suitable. The quality control of blood samples taken and absorbed on solid support is currently carried out by means of a visual examination by an operator.

Currently, therefore, the quality control systems and methods of biological samples and in particular of DBS are not automated and present the problem of having long times and reliability totally dependent on the accuracy of the operator. 3

A known solution for quality control is reported in patent US9150983 which describes a system and a method relating to the analysis of haematochemical data, or Dried- Blood-Spots (DBS), and having a first and a second station for detecting and processing the images of a sample. The first station has a first camera adjacent to a first surface of the sample slide and a first group of illumination directed to an opposite surface of the slide to capture the first image, while the second station has a second camera adjacent to the second surface and a second lighting group directed towards the first sliding surface to capture a second image. A computing device receives and analyzes the data of the detected images and identifies the position of the absorbed sample for removal using a punch. A second solution is reported in the patent CN107367622 which describes a completely automatic blood analysis system, which includes a host computer, a machine for transporting slides containing a biological sample, a storage and recognition block of Guthrie card, a system of perforation, an image acquisition and processing system, in which the host computer comprises a data storage subsystem, a display subsystem, and an image analysis and evaluation subsystem. 4

The processing device receives the image data detected in the first and second image evaluation stations and includes a series of instructions that analyze the data and identify a location of the absorbed sample for removal by means of a punch. According to the invention, the system aims to obtain the reuse of paper avoiding the waste of non-perforated parts and the automatic classification of samples.

A third solution CN107389681 describes an image processing method for automatic blood analyzes. The method comprises three main procedures, including industrial camera debugging, image acquisition and processing, and vision interpretation, in which the image acquisition and processing procedure includes four steps: performing photography using software, evaluating if the camera takes a photo successfully, evaluating if the photo should be repeated and assigning a unique identification code. The described method is applied for automatic systems for the analysis of biological samples, in particular of blood.

However, known systems do not allow the preliminary and automatic discrimination of biological samples suitable for examination, i.e. whose collection has been carried out correctly. In fact, the currently known quality control 5 systems and methods analyze an image to determine the point at which to punch the portion of paper to be analyzed, however they do not verify that the dry blood sample is in sufficient quantity, nor that the blood stain on the Guthrie card is centered and not smudged or with serum halos, i.e. that the sample is suitable for testing or whether it should be discarded and a new sample requested.

The purpose of the present invention is to provide a control quality system and method for biological samples, for example of DBS, capable of preliminary and automatically discriminating the samples suitable to be subjected to examination, therefore having characteristics such as to overcome the limits that still affect the current known systems. According to the present invention, control quality system for biological samples is provided, as defined in claim 1.

According to the present invention, a control quality method for biological samples is provided, as defined in claim 7.

For a better understanding of the present invention, a preferred embodiment is now described, purely by way of non- 6 limiting example, with reference to the attached drawings, in which:

Figure 1 shows a schematic diagram of a control quality system for biological samples in the process of analyzing biological samples, according to the invention;

Figure 2 shows a flow chart of a control quality method for biological samples, according to the invention.

With reference to these figures, and in particular to figure 1, a quality control system 100 of biological samples is shown, according to the invention.

In particular, the control quality system 100 for biological samples comprises: at least one acquisition block 101 for acquiring digital images relating to a biological sample, for example absorbed on a support, such as a Guthrie card, or contained in a test tube; at least one processing block 102 of the images acquired by the acquisition block 101;

- a processing system 103 comprising a digital archive 104 storing the images acquired by the acquisition block 101; 7 a dispatch center 110, connected to the processing block 102, to the acquisition block 101 and to the processing system 103.

According to an aspect of the invention, the acquisition block 101 is equipped with a data connection to a processing system 103.

According to an aspect of the invention, the image processing block 102 is based on artificial intelligence algorithms. According to an aspect of the invention, as shown in

Figure 1, the processing block 102, connected to a dispatch center 110, comprises a filter 102a suitable for a first check on the image acquired by the acquisition block 101, should the image file be damaged or not usable due to poor quality of the image itself, i.e. sample out of focus, dirty lens, sample image not detected, the image is automatically discarded by filter 102a and a signal is issued for a new image.

According to one aspect of the invention, the acquisition block 101 is a camera.

According to one aspect of the invention, the acquisition block 101 is a digital camera. 8

According to an aspect of the invention, the acquisition block 101 is a mobile device configured to take a photograph and send it to the processing system 103.

According to one aspect of the invention, the acquisition block 101 is a computer, or a tablet, or a smartphone including a camera or a means for acquiring a digital photographic image.

According to an aspect of the invention, the processing block 102 comprises a recognition module 102b designed to recognize elements such as prints, manuscripts, QR codes, signatures, etc. in the detected image. The recognition module 102b is also able to detect the type of biological sample declared corresponding for example to urine, blood, feces, etc. According to an aspect of the invention, block 102 further comprises a block 102c which implements artificial intelligence methods, based for example on machine learning algorithms, or deep learning, suitable for evaluating the quality of the detected image in terms of the position of the photographed sample, of size, shape, color, homogeneity, possible presence of halos, integrity of the support, etc. 9

This function is particularly suitable for evaluating the quality of biological samples.

According to an aspect of the invention, the processing block 102c is trained on the basis of data relating to the correctness of the sampling, such as: the image of the area covered with the sample, the homogeneity of deposition of the sample and the absence of halo and this processes the results relating to the images acquired by acquisition block 101.

Advantageously, the processing block 102 is adapted to establish the conformity of the sample and to be analyzed on the basis of the following parameters detected by the image: positioning, measurement and definition of the image of the biological sample, for example of the bloodstain in the DBS, by means of a mathematical recognition model, built and updated using artificial intelligence methods, for example machine learning, and connected to the reading unit to interpret the images of the sample, for example absorbed on the support, and acquired by the acquisition block 101.

The dispatch center 110 connected to the processing block 102, to the acquisition block 101 and to the processing system 103, is able to emit at least one output included in the group consisting of: 10

- Request for a new image, in the event of a negative result of the verification carried out by the filter 102a;

- Request for a new sample to be photographed, in the event of a negative outcome of the verification by filter 102a and / or by recognition module 102b;

Emission of images and results following image processing and sending to block 102c for training artificial intelligence methods;

- Output of images and results to a final recipient of the data, for example to a LIS system, or a management system of a hospital or an analysis laboratory.

The mathematical recognition model is refined as the database is updated with the results of the analyzes carried out by the processing block 102 in order to progressively increase the degree of accuracy of the system 100.

According to one aspect of the invention, the system 100 provides for the validation by an expert operator of the predictive results of the updated mathematical model before its implementation. According to an aspect of the invention, the processing block 102 is able to automatically read prints, handwritten annotations, or signs present on the sampling medium such as, 11 for example, preferably an Optical Character Recognition (OCR) system or similar system.

According to an aspect of the invention, the system 100 comprises a digital archive 104 for archiving images designed to provide training data for other artificial intelligence algorithms.

Furthermore, the digital archive 104 is suitable for verification by a user in the event that the system provides unclear results. Advantageously, according to the invention, the system

100 discriminates in a pre-analytical phase between samples suitable for the analytical phase and unsuitable samples, possibly requiring a second sampling of the latter.

Advantageously, according to the invention, the system 100 allows the automation of the pre-analytical quality control phases of biological samples entering a laboratory without the need for human intervention.

According to an aspect of the invention, the system 100 can be advantageously used both on the sampling site and in a centralized analysis laboratory where the separation between suitable and unsuitable samples can take place. 12

The system 100 according to the invention is able to send only the biological samples deemed compliant and suitable for analysis to subsequent stages of analysis.

According to an aspect of the invention, the dispatch center 110 is connected to a LIS (Laboratory Information System), for example of a hospital, allowing to integrate and synchronize the data relating to the biological samples processed by the system 100 with second data collected by a final recipient of the data. In use, the acquisition block 101 detects and acquires an image of the sample. A signal is sent from the acquisition block 101 to the dispatch center 110 to communicate the acquisition of a new image, from the dispatch center 110 the data is transmitted to the digital archive 104 and then to the processing block 102 for quality checks on the acquired image. Once the processing block 102 verifies the conformity of the sample, the data is transmitted from the dispatch center 110 to the laboratory or to an LIS system.

According to an aspect of the invention the system 100 comprises an I / 0 block configured to operate the acquisition block 101, this I / O block being able to emit a start pulse of the system 100 by means of for example a 13 software (LIS / LIMS) in an authorized analysis laboratory or through start-up by an operator.

As previously mentioned, the present invention also refers to a method for controlling the quality of biological samples.

In particular, the method 200 for controlling the quality of biological samples comprises the steps:

- 201 of biological sampling, that is the collection of a sample, for example a blood sample on a support, for example of absorbent paper for DBS;

- 202 of acquiring an image of the biological sample, by means of an acquisition block 101;

- 203 of saving the acquired image in a pre-established digital archive 104 of image; - 204 of establishing the conformity of the sample to be analyzed (by means of system 100 or other similar system) on the basis of: erection, measurement and definition of the biological sample stain on the support by means of a mathematical recognition model, constructed and updated using artificial intelligence methods and connected to the reading unit to interpret the images; 14

205 of sending only the biological samples deemed compliant in step 204 to subsequent steps of analysis;

206 of sending and updating all data acquired and processed, in real or deferred time, to information and communication systems of a health facility or to a final recipient.

Therefore, the system and the method for controlling the quality of biological samples according to the invention allow to discriminate the samples suitable for being analyzed automatically, quickly and accurately.

Furthermore, the system and the method for controlling the quality of biological samples according to the invention allow to evaluate if the sample absorbed on a support is in sufficient quantity, if it is centered, if it is homogeneous and if, in the case of DBS, it has sufficiently sharp edges to be able to be subjected to analysis.

Furthermore, the system and the method for controlling the quality of biological samples according to the invention allow to evaluate if the sample has been compromised due to biological processes, in the case of test-tube samples if it has undergone haemolysis, in the case of DBS if the sample 15 has undergone separation of the serous component from the other blood components that can be subjected to analysis.

Finally, the system and the method for controlling the quality of biological samples according to the invention allow to increase the reliability of quality control using artificial intelligence.

Finally, it is clear that modifications and variations may be made to the system and the method for controlling the quality of biological samples described and illustrated here without departing from the protective scope of the present invention, as defined in the attached claims.