AI helps predict lung cancer severity

A computer program based on data from nearly a half-million tissue images and powered by artificial intelligence (AI) can accurately diagnose in cases of adenocarcinoma, the most common form of lung cancer. Researchers at New York University Langone Health’s Perlmutter Cancer Center and the University of Glasgow, in Scotland, developed and tested the program. 

The algorithm, or more specifically histomorphological phenotype learning (HPL), was found to accurately distinguish between similar lung cancers – adenocarcinoma and squamous cell cancers – 99% of the time. The HPL program was also found to be 72% accurate at predicting the likelihood and timing of cancer’s return after therapy, bettering the 64% accuracy in the predictions made by pathologists who directly examined the same patients’ tumour images, researchers said.

The research team also points out that the program is independent and “self-taught,” meaning that it determined on its own which structural features were statistically most significant for gauging the severity of disease and had the greatest impact on tumour recurrence. They stress that because HPL is self-learning, the program will become increasingly more accurate as more data is added over time. To build public trust, researchers have posted their programming code online and have plans to make the new HPL tool freely available on completion of further testing. The study was published recently in Nature Communication.

Oral rinse test for gastric cancer

A simple oral rinse may lead to early detection of gastric cancer, according to new findings presented recently at ASCO’s Annual Meeting 2024 in Chicago, USA, and simultaneously published in Gastroenterology. In the new study, researchers analysed bacteria samples from the mouths of 98 patients scheduled to undergo endoscopy – 30 of whom were known to have gastric cancer, 30 of whom had premalignant gastric conditions, and 38 of whom were healthy controls. 

The researchers found distinct differences between the oral microbiomes of the controls compared with patients with gastric cancer and premalignant conditions. They also discovered little difference between the samples collected from the patients with gastric cancer and those with premalignant conditions, suggesting that changes in the microbiome may occur as soon as the stomach environment begins to undergo changes that can eventually develop into cancer.

The findings indicated that oral bacteria alone could serve as biomarkers for the risk of gastric cancer. After conducting their research, the researchers developed a model of the 13 bacterial genera representing the most significant differences between controls and the gastric cancer and premalignant conditions groups. They plan to conduct larger studies involving multiple institutions to ensure that their findings are generalisable to a wider population.

AI tech may help cut cancer waiting lists

Cancer waiting times are set to fall thanks to new AI technology that locates cancer cells 2.5 times quicker than doctors alone. Game-changing AI is being rolled out to every NHS radiotherapy department in England – backed by £15.5 million in new Government funding.

It works by automatically reviewing a CT or MRI scan, helping doctors quickly distinguish between cancerous cells and healthy organs and to prevent healthy organs from being damaged during radiation treatment. Trained health workers will of course review any report before administering any treatment – helping tens of thousands of cancer patients each year get faster treatment.

Dr Imogen Locke, national speciality adviser for radiotherapy at NHS England said: “The NHS is embracing AI and its benefits for cancer patients and every radiotherapy department will soon be able to offer the latest technology to help diagnose and treat patients more quickly. We are seeing a record number of referrals for suspected cancer, and game-changing tools like AI will help the NHS continue the significant progress made in tackling the longest waits for patients.”

e-nose tech in early-stage lung cancer

Electronic nose (e-nose) technology identified early-stage lung cancer with high reliability in a prospective observational clinical trial conducted at Memorial Sloan Kettering Cancer Center (MSK) in the US. 

E-nose diagnostic predictions agreed with histopathological results in 86% of 100 patients with clinical stage 1 lung cancer, according to the study published recently in the Journal of Thoracic Oncology. “These robust results demonstrate e-nose technology could be a reliable complement to existing diagnostic methods for early-stage lung cancer,” said principal investigator MSK thoracic surgeon Gaetano Rocco. 

“We anticipate this non-invasive, low-cost, portable, and highly reliable diagnostic method will revolutionise the diagnosis and clinical management of early-stage lung cancer, making it accessible to more patients, especially those resistant to CT scans or for whom a biopsy is infeasible. Our immediate next step is applying for funding to support our efforts to advance research and development.”

The present MSK-exclusive trial included 100 treatment-naïve adult patients ages 21 to 85 evaluated at MSK between 2020 and 2023. E-nose results agreed with histopathologic results for 86% of 100 cases analysed. The device earned an F1 score of 92.5% based on 86 true positives, two false negatives, and 12 false positives. Dr Rocco said the next steps will include refining the device from its current size of about three feet long to a smaller scale, perhaps as small as a USB flash drive fitting easily in a clinician’s jacket pocket – this miniaturisation process should take just a few months.