The use of AI to streamline drug discovery is exploding. Researchers are deploying machine-learning models to help them identify molecules, among billions of options, that might have the properties they are seeking to develop new medicines.
But there are so many variables to consider—from the price of materials to the risk of something going wrong—that even when scientists use AI, weighing the costs of synthesizing the best candidates is no easy task.
The myriad challenges involved in identifying the best and most cost-efficient molecules to test is one reason new medicines take so long to develop, as well as a key driver of high prescription drug prices.
To help scientists make cost-aware choices, MIT researchers have developed an algorithmic framework to automatically identify optimal molecular candidates, which minimizes synthetic cost while maximizing the likelihood candidates have desired properties. The algorithm also identifies the materials and experimental steps needed to synthesize these molecules.
In a study published in Nature Medicine, an interdisciplinary team of researchers from The Westmead Institute for Medical Research (WIMR) and the Sydney Precision Data Science Centre at the University of Sydney have, for the first time, identified molecular biomarkers for transplant rejection that are common to all the major transplanted organs: hearts, lungs, livers, and kidneys.
This significant advancement led by Harry Robertson, a Ph.D. student at the University of Sydney, and aspiring bioinformatician at WIMR, uses machine learning to predict transplant outcomes with unprecedented accuracy.
Mr. Robertson said, "Our research indicates that there are underlying molecular pathways involved in organ rejection that are consistent across different solid organs. This discovery is pivotal as it allows us to develop strategies to enhance the success rates of all transplants."
Working together, the team created the Pan-organ ResOurce for Molecular Allograft Dysfunction (PROMAD), a molecular atlas consisting of more than 12,000 patient samples from around the globe. This atlas will enable researchers to interact with data on transplants that was previously inaccessible to many researchers, and encourage global collaboration.
A UCLA-led team has developed a machine-learning model that can predict with a high degree of accuracy the short-term survival of dialysis patients on continuous renal replacement therapy (CRRT). The study is published in Nature Communications.
CRRT is a therapy used for very sick hospitalized patients whose health status makes them ineligible for regular hemodialysis. It is a gentler therapy that provides continuous treatment over a prolonged period. About half of adults placed on CRRT, however, do not survive, rendering the treatment futile for both patients and their families.
A new study has found that an artificial intelligence (AI) conversational agent enhances patient care after cataract surgery. The AI-powered automated voice system, called Dora, is able to call patients to ask them questions, understand their answers and accurately identify responses that indicate the need for clinical review.
The study, conducted by researchers at Newcastle University, Oxford University Hospitals (OUH) NHS Foundation Trust and Imperial College Health care NHS Trust, was published in the journal eClinicalMedicine.
Combining the principles of evolution with artificial intelligence (AI), scientists have proposed a new way to predict the chance of prostate cancer returning. In a recent study, they harnessed computational methods to capture specific tumor measurements relating to the tumor's ability to change over time. They then showed that these measurements correlate with disease recurrence more than a decade after the initial diagnosis.
This approach could help clinicians systematically categorize patients according to their risk of disease recurrence. Based on this, they may be able to determine which patients only need localized treatment—typically radiotherapy, often alongside hormone therapy, or surgery—and which should receive additional treatment.
Developing a new drug can take years of research and cost millions of dollars. Still, more than 90% of drug candidates fail in clinical trials, with even more that never make it to the clinical stage. Many drugs fail because they simply aren't safe.
Researchers at the Broad Institute of MIT and Harvard have developed AI models that can screen the potential biological effects of drugs before they ever enter a living organism.
That's a significant breakthrough. The AI models developed by the researchers at the Broad Institute of MIT and Harvard can potentially streamline the drug development process and reduce the risk of adverse reactions. By screening the potential biological effects of drugs before they're tested in living organisms, the AI models can help identify potential issues earlier on, saving time and resources. This could also lead to the development of safer and more effective treatments.
A new machine learning model can predict autism in young children from relatively limited information. This is shown in a new study by Karolinska Institutet published in JAMA Network Open. The model can facilitate early detection of autism, which is important to provide the right support.
"With an accuracy of almost 80% for children under the age of two, we hope that this will be a valuable tool for health care," says Kristiina Tammimies, Associate Professor at KIND, the Department of Women's and Children's Health, Karolinska Institutet and last author of the study.
The research team used a large US database (SPARK) with information on approximately 30,000 individuals with and without autism spectrum disorders.
Researchers have developed an artificial intelligence which can differentiate cancer cells from normal cells, as well as detect the very early stages of viral infection inside cells. The findings, published today in a study in the journal Nature Machine Intelligence, pave the way for improved diagnostic techniques and new monitoring strategies for disease. The researchers are from the Centre for Genomic Regulation (CRG), the University of the Basque Country (UPV/EHU), Donostia International Physics Center (DIPC) and the Fundación Biofisica Bizkaia (FBB, located in Biofisika Institute).
The tool, AINU (AI of the NUcleus), scans high-resolution images of cells. The images are obtained with a special microscopy technique called STORM, which creates a picture that captures many finer details than what regular microscopes can see. The high-definition snapshots reveal structures at nanoscale resolution.
Head and neck cancers have increased significantly over the last 30 years. In Germany, there are about 18,000 to 20,000 new cases of head and neck tumors every year. In particular, the incidence of carcinomas of the middle pharynx has increased, which is associated with the increase in human papillomavirus (HPV) infections.
Using a machine-learning-based method, an interdisciplinary team of researchers led by Sara Wickström at the University of Helsinki, in collaboration with the University of Turku and the Max Planck Institute for Molecular Biomedicine in Germany, has analyzed hundreds of biobank patient samples at the level of individual cells. The new technology combines indicators of cancer cell behavior and the architecture of the tumor and surrounding healthy tissue to create a kind of "fingerprint" for each patient that can be used to assess prognosis and response to cancer therapy.
In another triumph for AI in healthcare, researchers have developed a model that can spot bits of brain tumors that surgeons may miss while removing them from patients. It can detect these remaining tissues in as little as 10 seconds, and help prevent a host of long- and short-term post-procedure complications.
Developed by University of Michigan and University of California San Francisco researchers, the technology is called FastGlioma – incorporating the term 'glioma' that refers to a brain or spinal cord tumor.
"The technology works faster and more accurately than current standard of care methods for tumor detection and could be generalized to other pediatric and adult brain tumor diagnoses," said neurosurgeon Todd Hollon, a senior author of the paper detailing FastGlioma's effectiveness that appeared in Nature. "It could serve as a foundational model for guiding brain tumor surgery.”
With most tumor removal surgeries, it's difficult to tell healthy brain tissue and tumorous tissue apart – and as a result, a bit of residual tumor could remain in the cavity from where the mass was removed.
Researchers at Karolinska Institutet have investigated how well different AI models can predict the prognosis of triple-negative breast cancer by analyzing certain immune cells inside the tumor. The study, published in the journal eClinicalMedicine, is an important step toward using AI in cancer care to improve patient health.
Tumor-infiltrating lymphocytes are a type of immune cell that plays an important role in fighting cancer. When they are present in a tumor, it means that the immune system is trying to attack and destroy the cancer cells.
These immune cells can be important in predicting how a patient with so-called triple-negative breast cancer will respond to treatment and how the disease will progress. But when pathologists assess the immune cells, the results can vary. Artificial intelligence (AI) can help standardize and automate this process, but it has been difficult to demonstrate that AI works well enough to be used in health care.
An observational, multicenter, real-world study conducted at 12 screening sites in Germany has reported a 17.6% higher cancer detection rate among women aged 50–69 who received AI-supported double-reading mammography screenings compared to those who received standard double-reading. Recall rates remained unchanged.
Mammography screening programs often rely on double reading to identify breast cancer at earlier stages. Radiologists face substantial workloads interpreting mammograms, most of which include cases with no signs of cancer. Screening centers struggle to keep up with providing efficient and accurate assessments, a problem only getting more urgent with a growing shortage of trained radiologists.
Many breast cancers elude early detection only to be diagnosed at later stages, reflecting ongoing issues with current screening methods. False positive results burden both participants and health care systems with needless worry and unnecessary follow-up (recall) appointments. Efforts to boost early detection sensitivity and lower unnecessary false positives are top priorities.
Stanford Medicine's AI Model Accurately Predicts Cancer Prognoses, Treatment Efficacy
The model is the first of its kind to use multiple types of imaging and language-based data to assess a cancer patient's health.
By Adrianna Nine January 15, 2025 https://www.extremetech.com/science/sta ... -treatment
Stanford Medicine has developed an artificial intelligence model that can accurately predict cancer patients' prognoses and responses to treatment. The first of its kind to leverage multiple types of imaging and language-based data, the model has already shown promise with several forms of cancer, including lung cancer, gastroesophageal cancer, and melanoma.
Over the last few years, researchers have created a range of experimental AI models that examine imaging data for tiny signs of cancer that doctors and radiologists might easily miss. Early tests show that these models are highly effective. Sybil, a model developed by MIT and the Massachusetts General Cancer Center, can predict patients' one-year lung cancer development with an 86% to 94% accuracy rate, while Harvard Medical School's pancreatic cancer prediction model can map a patient's three-year prognosis with 88% accuracy. Another MIT model even spots signs of the riskiest forms of breast cancer to shield patients from overtreatment.
New AI picks up 97% of lung diseases, and can tell pneumonia from COVID-19
By Bronwyn Thompson
January 23, 2025
A breakthrough new AI model is able to detect the presence of different lung diseases from ultrasound videos, with 96.57% accuracy, and it is even able to distinguish whether the abnormalities are due to pneumonia, COVID-19 or other conditions.
The model, developed by researchers at Australia's Charles Darwin University (CDU), United International University and the Australian Catholic University (ACU), can identify specific patterns of different lung disease, outperforming previous AI tools that have been tested on the same ultrasound datasets.
“The model also uses AI techniques to show radiologists why it made certain decisions, making it easier for them to trust and understand the results,” said study co-author Niusha Shafiabady, a professor at CDU. “This model helps doctors diagnose lung diseases quickly and accurately, supports their decision-making, saves time, and serves as a valuable training tool.”
Diagnostic AI Model Can Distinguish Between Multiple Types of Lung Disease
LungNet can reportedly tell the difference between pneumonia, COVID-19, and the flu.
By Adrianna Nine January 27, 2025 https://www.extremetech.com/science/dia ... ng-disease
Researchers from Australia and Bangladesh have developed an artificial intelligence model capable of spotting and distinguishing between multiple types of lung disease. Called LungNet, the tool already outperforms other lung-focused AI models and can justify its "decisions," making it a potentially useful—and life-saving—companion to human medical experts.
Of all the verticals that stand to benefit from analytical AI, the medical field is near the top of the list. Researchers from a number of specialties are busy designing AI models that can identify signs of cancer from X-rays, microscope slides, and CT scans. Some even predict patients' prognoses and inform treatment plans. But most of these models tend to focus on ruling out a single condition, like lung cancer or heart disease. Very few can look at a part of the body and hunt for the multiple types of illnesses that may lurk there.
Researchers at the University of Jyväskylä, in collaboration with the University of Turku's Institute of Biomedicine, University of Helsinki and Nova Hospital of Central Finland, have developed an advanced artificial intelligence tool for automatic analysis of colorectal cancer tissue slides.
The neural network model developed in the study outperformed all previous models in the classification of tissue microscopy samples. The research is published in the journal Heliyon.
"Based on our study, the developed model is able to identify all tissue categories relevant for cancer identification, with an accuracy of 96.74%," Fabi Prezja, the researcher responsible for the design of the method, says.
In what's expected to soon be commonplace, artificial intelligence is being harnessed to pick up signs of cancer more accurately than the trained human eye. This latest AI model has a near 100% success rate and serves as a clear sign of things to come.
An international team of scientists including those from Australia's Charles Darwin University (CDU) has developed a novel AI model known as ECgMPL, which can assess microscopic images of cells and tissue to identify endometrial cancer – one of the most common forms of reproductive tumors – with an impressive 99.26% accuracy. And the researchers say it can be adapted to identify a broad range of disease, including colorectal and oral cancer.
“The proposed ECgMLP model outperforms existing methods by achieving 99.26 percent accuracy, surpassing transfer learning and custom models discussed in the research while being computationally efficient,” said the study's co-author Dr. Asif Karim, from CDU. “Optimized through ablation studies, self-attention mechanisms, and efficient training, ECgMLP generalizes well across multiple histopathology datasets thereby making it a robust and clinically applicable solution for endometrial cancer diagnosis.”
Patients with narrowing of at least 50% in three major coronary arteries did equally well when treated with a minimally invasive stent placement guided either by ultrasound-based imaging or by a novel, artificial-intelligence-powered (AI), non-invasive imaging technique derived from angiography, researchers reported at the American College of Cardiology's Annual Scientific Session (ACC.25) on March 30 in Chicago. The work was simultaneously published in The Lancet.
"This is the first such study to be conducted in patients with angiographically significant lesions," said Jian'an Wang, MD, a professor in the Heart Center at The Second Affiliated Hospital of Zhejiang University School of Medicine in Hangzhou, China, and the study's senior author. "Patients whose evaluation was non-invasively guided by the novel, AI-powered technique underwent approximately 10% fewer procedures, and their outcomes were comparable with those for patients whose evaluation was guided by a commonly used ultrasound-based imaging technique."