![Four state-of-the art image search engines fail for routine clinical care in histopathology – UCLA Health](https://oncodaily.com/pub/uploads/2024/05/download-49.jpg)
Four state-of-the art image search engines fail for routine clinical care in histopathology – UCLA Health
Four proposed state-of-the art image search engines for automating search and retrieval of digital histopathology slides were found to be of inadequate performance for routine clinical care, new research suggests.
Dr. Helen Shang, a third-year internal medicine resident and incoming hematology-oncology fellow at the David Geffen School of Medicine at UCLA, co-led the study with Dr. Mohammad Sadegh Nasr of the University of Texas at Arlington.
According to Dr Shang, the performance of the artificial intelligence algorithms to power the histopathology image databases was worse than expected, with some having less than 50% accuracy, which is not suitable for clinical practice.
“Currently, there are many AI algorithms being developed for medical tasks but there are fewer efforts directed on rigorous, external validations. The field has also yet to standardize how AI algorithms should be best tested prior to clinical adoption.”, – said Shang.
The paper is published in the peer-reviewed journal NEJM AI.
As it now stands, pathologists manually search and retrieve histopathology images, which is very time consuming. As result, there has been growing interest in developing automated search and retrieval systems for the digitized cancer images.
The researchers designed a series of experiments to evaluate the accuracy of search engine results on tissue and subtype retrieval tasks on real-world UCLA cases and larger, unseen datasets. The four engines examined are Yottixel, SISH, RetCCL, HSHR. Each takes a different approach toward indexing, database generation, ranking and retrieval of images.
Overall, the researchers found inconsistent results across the four algorithms – for instance, Yottixel performed best on breast tissue, while RetCCL had the highest performance on brain tissue. They also found that a group of pathologists found search engine results to be of low to average quality with several visible errors.
The researchers are devising new guidelines to standardize the clinical validation of AI tools, Shang said. They are also developing new algorithms that leverage a variety of different data types to develop more reliable and accurate predictions.
“Our studies show that despite amazing progress in artificial intelligence over the past decade, significant improvements are still needed prior to widespread uptake in medicine. These improvements are essential in order to avoid doing patients harm while maximizing the benefits of artificial intelligence to society.”, – Shang said.
The study was funded by the University of Texas System Rising STARs Award and the CPRIT First Time Faculty Award.
Histopathology Slide Indexing and Search — Are We There Yet?
Authors: Helen Shang, Mohammad Sadegh Nasr, Jai Prakash Veerla, Jillur Rahman Saurav, Amir Hajighasemi, Parisa Malidarreh, Manfred Huber, Chace Moleta, Jitin Makker, and Jacob Luber.
Source: UCLA Health
UCLA Health is the public healthcare system affiliated with the University of California, Los Angeles, located in Los Angeles, California. With world-renowned facilities like the Ronald Reagan UCLA Medical Center and a network of specialized clinics, UCLA Health consistently ranks among the top medical centers worldwide. It remains dedicated to advancing biomedical research and fostering innovations in areas such as organ transplantation, neurosurgery, and cancer treatment.
Helen Shang is a third-year internal medicine resident and incoming hematology-oncology fellow at the David Geffen School of Medicine at UCLA and a Visiting Assistant Professor of Computer Science at the University of Texas at Arlington.
Mohammad S. Nasr is a researcher based at the Luber Research Group, University of Texas, Arlington, with a focus on the application of artificial intelligence (AI) and machine learning (ML) in the field of cancer research. His specific areas of expertise include enhancing the interpretability of medical models, investigating deep Bayesian networks, and advancing drug discovery techniques.
The Luber Research Group at The University of Texas at Arlington focuses on computational tools for cancer imaging, aiming to improve patient care and drug discovery. Their expertise includes handling high throughput ‘omics data and building interpretable deep Bayesian models relevant for clinical use. Additionally, they investigate the microbiome’s impact on patient response to cancer immunotherapy. Supported by prestigious grants, including from the Cancer Prevention and Research Institute of Texas, the lab is dedicated to advancing cancer research and therapy.