AI Revolutionizes Breast Cancer Diagnosis and Staging

AI improves breast cancer staging by analyzing chromatin images

Paul Scherrer Institut (PSI)Jul 24 2024 Researchers at the Paul Scherrer Institute PSI and the Massachusetts Institute of Technology MIT are using artificial intelligence to improve the categorization of breast cancer. Not all cancers are the same. Some tumors grow very slowly or hardly ever change from a comparatively harmless pre-cancerous form to a life-threatening form. In men, this includes prostate cancer and, in women, a precursor of breast cancer in the milk ducts, known as ductal carcinoma in situ. In 30 to 50 percent of cases, this form, abbreviated to DCIS, develops into a threatening invasive breast carcinoma. Because DCIS is highly curable, doctors generally recommend treatment. Until now, doctors have lacked the necessary indicators to reliably decide which tumors will remain benign and which will become a life-threatening invasive ductal carcinoma (IDC). This lack of knowledge in characterizing breast cancer prompted a new study, led by G.V. Shivashankar, Head of the Laboratory of Nanoscale Biology at PSI and Professor of Mechano-Genetics at ETH Zurich, and Caroline Uhler, Director of the Eric and Wendy Schmidt Centre at the Broad Institute and Professor of Electrical Engineering and Computer Science at MIT. The researchers have developed a system for analyzing images which uses artificial intelligence to reliably determine the stage of the disease. "Our work opens up a unique approach to identifying the stage of DCIS using images that show how the DNA is packaged in each individual cell. Collecting this data is simple and inexpensive," explains Shivashankar. Women live with uncertainty when making treatment decisions DCIS accounts for about 25 percent of all breast cancer diagnoses. The cells lining a patient's milk ducts look different from healthy tissue, and often microcalcifications are visible. Treatment can take the form of radiotherapy, hormone therapy or surgery. In clinical practice, doctors use a process known as grading to determine the prognosis for DCIS and select a suitable therapy. This involves classifying the amount of change and assigning the result to one of seven different categories. These describe features such as the size of the DCIS, the appearance of the cell nuclei, whether it has grown (hyperplasia), whether the cells have entered neighboring tissue (invasive), whether they have spread into lymph or blood cells (aggressive) or whether they are in the process of forming secondary tumors (metastatic). However, the progression from DCIS to a serious form of IDC is by no means a certainty - 50 to 70 percent of cases remain benign. But which? Scientists are pursuing different approaches to make their forecasts more reliable. For example, sophisticated imaging technology is being used to identify indicators of the risk posed by an early form of the disease. Another approach involves extensive transcriptome analyses. These use sequencing to determine how many and which genes are active in suspect cells at a certain point in time. However, these approaches have not yet been tested in everyday clinical practice, and they are too complicated and too expensive to be practicable. For the women concerned, deciding on the right treatment remains fraught with uncertainty: they face the prospect of undergoing treatment that may not only be unnecessary, but could also harbour the risk of side effects. AI improves DCIS staging The current study shows that artificial intelligence (AI) can improve staging using data that is easy and inexpensive to collect. The researchers, led by Shivashankar and Uhler, trained a machine learning algorithm on 560 tissue samples from 122 patients. These had been stained with the dye DAPI, which makes the chromatin in the cell's nucleus fluoresce. Chromatin consists, among other things, of DNA and proteins. Based on its appearance, conclusions can be drawn about the organization and thus the activity of the DNA in the cell nucleus. After a learning phase, the AI model was able to identify patterns in the tissue sections that matched the differences identified by human pathologists. "Our analysis shows that chromatin images, which are cheap and easy to obtain, together with powerful AI algorithms, can provide enough information to study how the cell state and tissue organization change during the transition from DCIS to IDC, and thereby accurately predict the stage of the disease," explains Uhler. The researchers believe that this kind of tumor classification based on AI and chromatin imaging has great potential. However, before it can be used in practical applications, numerous further studies are needed to demonstrate the reliability and safety of the approach, including long-term monitoring of DCIS patients. Paul Scherrer Institut (PSI) Journal reference: Zhang, X., et al. (2024). Unsupervised representation learning of chromatin images identifies changes in cell state and tissue organization in DCIS. Nature Communications. doi.org/10.1038/s41467-024-50285-1.

AI model enhances DCIS diagnosis using tissue image analysis

Massachusetts Institute of TechnologyJul 24 2024 Ductal carcinoma in situ (DCIS) is a type of preinvasive tumor that sometimes progresses to a highly deadly form of breast cancer. It accounts for about 25 percent of all breast cancer diagnoses. Because it is difficult for clinicians to determine the type and stage of DCIS, patients with DCIS are often overtreated. To address this, an interdisciplinary team of researchers from MIT and ETH Zurich developed an AI model that can identify the different stages of DCIS from a cheap and easy-to-obtain breast tissue image. Their model shows that both the state and arrangement of cells in a tissue sample are important for determining the stage of DCIS. Because such tissue images are so easy to obtain, the researchers were able to build one of the largest datasets of its kind, which they used to train and test their model. When they compared its predictions to conclusions of a pathologist, they found clear agreement in many instances. In the future, the model could be used as a tool to help clinicians streamline the diagnosis of simpler cases without the need for labor-intensive tests, giving them more time to evaluate cases where it is less clear if DCIS will become invasive. "We took the first step in understanding that we should be looking at the spatial organization of cells when diagnosing DCIS, and now we have developed a technique that is scalable. From here, we really need a prospective study. Working with a hospital and getting this all the way to the clinic will be an important step forward," says Caroline Uhler, a professor in the Department of Electrical Engineering and Computer Science (EECS) and the Institute for Data, Systems, and Society (IDSS), who is also director of the Eric and Wendy Schmidt Center at the Broad Institute of MIT and Harvard and a researcher at MIT's Laboratory for Information and Decision Systems (LIDS). Uhler, co-corresponding author of a paper on this research, is joined by lead author Xinyi Zhang, a graduate student in EECS and the Eric and Wendy Schmidt Center; co-corresponding author GV Shivashankar, professor of mechogenomics at ETH Zurich jointly with the Paul Scherrer Institute; and others at MIT, ETH Zurich, and the University of Palermo in Italy. The open-access research was published in Nature Communications. Combining imaging with AI Between 30 and 50 percent of patients with DCIS develop a highly invasive stage of cancer, but researchers don't know the biomarkers that could tell a clinician which tumors will progress. Researchers can use techniques like multiplexed staining or single-cell RNA sequencing to determine the stage of DCIS in tissue samples. However, these tests are too expensive to be performed widely, Shivashankar explains. In previous work, these researchers showed that a cheap imagining technique known as chromatin staining could be as informative as the much costlier single-cell RNA sequencing. For this research, they hypothesized that combining this single stain with a carefully designed machine-learning model could provide the same information about cancer stage as costlier techniques. First, they created a dataset containing 560 tissue sample images from 122 patients at three different stages of disease. They used this dataset to train an AI model that learns a representation of the state of each cell in a tissue sample image, which it uses to infer the stage of a patient's cancer. However, not every cell is indicative of cancer, so the researchers had to aggregate them in a meaningful way. They designed the model to create clusters of cells in similar states, identifying eight states that are important markers of DCIS. Some cell states are more indicative of invasive cancer than others. The model determines the proportion of cells in each state in a tissue sample. Organization matters "But in cancer, the organization of cells also changes. We found that just having the proportions of cells in every state is not enough. You also need to understand how the cells are organized," says Shivashankar. With this insight, they designed the model to consider proportion and arrangement of cell states, which significantly boosted its accuracy. The interesting thing for us was seeing how much spatial organization matters. Previous studies had shown that cells which are close to the breast duct are important. But it is also important to consider which cells are close to which other cells." Xinyi Zhang, lead author When they compared the results of their model with samples evaluated by a pathologist, it had clear agreement in many instances. In cases that were not as clear-cut, the model could provide information about features in a tissue sample, like the organization of cells, that a pathologist could use in decision-making. This versatile model could also be adapted for use in other types of cancer, or even neurodegenerative conditions, which is one area the researchers are also currently exploring. "We have shown that, with the right AI techniques, this simple stain can be very powerful. There is still much more research to do, but we need to take the organization of cells into account in more of our studies," Uhler says. This research was funded, in part, by the Eric and Wendy Schmidt Center at the Broad Institute, ETH Zurich, the Paul Scherrer Institute, the Swiss National Science Foundation, the U.S. National Institutes of Health, the U.S. Office of Naval Research, the MIT Jameel Clinic for Machine Learning and Health, the MIT-IBM Watson AI Lab, and a Simons Investigator Award. Massachusetts Institute of Technology Journal reference: Zhang, X., et al. (2024). Unsupervised representation learning of chromatin images identifies changes in cell state and tissue organization in DCIS. Nature Communications. doi.org/10.1038/s41467-024-50285-1.