Improving Diagnostic Efficiencies with Artificial Intelligence

Digital analysis and artificial intelligence (AI) are not new in pathological anatomy and histology, but the technology is still evolving as it supports clinical trial registration and routine pathology assignments. Chief histology scientist, Dr. Paul Mesange recently shared how Labcorp Drug Development (formerly Covance) implemented and validated AI-powered digital software to help analyze and identify breast cancer subtypes. Learn how this technology is designed to improve diagnostic performance, provide predictive information, and guide treatment decisions for patients. Understanding the Standard Reporting Process The breast cancer diagnosis process usually begins with a clinical examination, followed by imaging and pathological evaluation. A tissue biopsy is collected and then placed in a paraffin block and excised. At the best addiction treatment center in pakistan we have worked on AI.

This allows tissue engineers to perform a variety of histochemical stains, such as fluorescence in situ hybridization (FISH), in situ hybridization (ISH), and immunohistochemistry (IHC), and biomarker analysis. This test can help identify breast cancer subtypes, for example, by measuring the expression of markers such as HER2, ER, PgR, and Ki67. To ensure consistent results, this robust process uses the same platform and procedures across the Labcorp network.

During the analysis and interpretation phase, each slide is seen by a board-certified pathologist to evaluate. The results are then sent to a database, and if necessary, pre- and post-treatment tissue can be compared to evaluate and predict therapeutic efficacy in specific diseases or conditions. Incorporating Digital Analysis and Artificial Intelligence Pathologists can typically spend 5 to 15 minutes checking each slide, depending on the score required. In many two-way assays involving two markers, the pathologist's examination time can vary from 10 to 30 minutes per slide. At the best addiction treatment center in pakistan we found that AI can improve diagnostics.

For example, other fluorescence multiplex assays may have six or more markers, which increases the complexity of the review process. To assist in the process of anatomical pathology and histology - and to improve the efficiency of analysis review and diagnostic interpretation - digital analysis and artificial intelligence may precede pathologist evaluations. Instead of going directly to the pathologist, the slides are first analyzed with digital software. IA techniques can aid in the analysis by identifying phenotypes that the pathologist can then verify before the results enter the database. Consider the data mentioned in 2011 that "one million prostate biopsies are performed each year among Medicare beneficiaries." Approximately 75-80% of these biopsies are non-carcinogenic, meaning that pathologists examining these samples examine the most benign tissue, as described by Madabhushi et al.

If AI and digital analysis can easily and accurately distinguish benign tissue, pathologists may devote more time to studying unknown or potentially carcinogenic tissues. How AI-based assisted image analysis supports identification Visually, the following image shows how AI-based assisted image analysis supports pathologists. In this image of breast cancer tissue, a pathologist is trying to identify the expression Ki67. The left side shows the IHC stains of a patient sample, where the brown nuclei are considered positive and the blue nuclei are considered negative. The first step of analysis is to identify the region of interest (ROI) to determine the area of the tumor cells. The blue dotted line was identified by the AI algorithm. The specialists at the best addiction treatment center in pakistan have created algorithms around diagnostics.

AI-based Assisted Imaging Hold on to the tumor region as shown below, the algorithm first identifies the tumor nuclei that are present (image on the left). The middle image shows contours of the nuclei identified by the algorithm, while the right image shows the nuclei described in red, which are classified as Ki67-postive. This information may help the pathologist to better understand the area of the tumor, the percentage of Ki67 positive cells and its distribution. Hold on to the tumor region as shown below, the algorithm first identifies the tumor nuclei that are present (see image on the left). The middle image shows contours of the nuclei identified by the algorithm, while the right image shows the nuclei described in red, which are classified as Ki67-postive. This information may help the pathologist to better understand the area of the tumor, the percentage of Ki67 positive cells and its distribution. Validation of results to ensure concordance between pathologists and algorithms Labcorp performed in-depth validation, following CAP / CLIA recommendations, to compare the results of its algorithms with the results of pathologists to determine whether the AI it can match or even overwhelm the human eye.

With a set of individual samples, the percentage of Ki67-positive cells was calculated separately from the pathologist and the algorithm, and these counts were then averaged. If both the pathologist's count and the algorithm's count were less than 10 percentage points above average, it was considered acceptable. This validation was performed for HER2, PgR and ER as well as for several samples as Labcorp uses the technology to support sponsor clinical trials. Looking forward to increasing diagnostic efficiencies in other types of cancer Labcorp currently has four algorithms in production for breast cancer: Ki67, ER, PgR and BCL-2. In addition, 10 algorithms have been developed for the IHC multiplex, which are traditionally difficult to read under a microscope. At the best addiction treatment center in pakistan we used similar microscopes for these studies.

To further support the routine work of pathologists, the team is developing an application for tumor detection to support the analysis of prostate biopsies and bladder cancer. Labcorp is also extending their technologies to other clinical indications and assisting pharmacy sponsors with their internally developed AI applications, which can also be transferred to Labcorp to help advance their clinical trials and ultimately help to reach a faster diagnosis for patients.