Call for Papers: Bio-inspired Deep Learning Image and Signal Processing Pipelines in Medical Oncology
PeerJ has recently launched a new Collection on Bio-inspired Deep Learning Image and Signal Processing Pipelines in Medical Oncology: From Radiomics to Pathomics for non-invasive oncological early diagnosis and outcome prediction. This collection will look to bring together interesting and relevant peer-reviewed articles published in PeerJ Computer Science and PeerJ – the Journal of Life and Environmental Sciences on deep learning methodologies in medical imaging, specifically looking at how to improve accurate diagnosis and precise treatment for cancer patients.
Computer science approaches to artificial intelligence and advanced mathematical modeling are becoming increasingly relevant for medical diagnosis. This collection offers a space to look across different approaches in oncology and encourages studies that outline the practical implementation of these pipelines in clinical settings.
This collection is co-organized by Dr. Eng. Francesco Rundo (STMicroelectronics), Dr. Giuseppe Luigi Banna (Department of Medical Oncology – United Lincolnshire Hospitals NHS Trust, Lincoln, UK), Prof. Concetto Spampinato (University of Catania) and Prof. Sabrina Conoci (University of Messina). We asked the organizers a few questions to get a better understanding of what kind of studies this Collection is looking to highlight.
What sparked the idea for a collection on this topic?
The encouraging results that we have obtained by applying modern methods of artificial intelligence to the field of medical oncology have led us to form this cross-disciplinary group that is setting itself the ambitious goal of stimulating researchers from various areas to investigate these new advanced approaches to improve the health of cancer patients as well as their prognosis with an accurate early diagnosis.
What are you looking to achieve with bringing together papers on this topic?
Our goal, as mentioned, is to improve the health of cancer patients by means of an accurate diagnosis and predictive estimation of the response to drug treatment. Furthermore, we believe that these approaches can also improve the fight against tumors because they help facilitate early diagnoses using non-invasive methods and at sustainable costs.
How is deep learning for medical imaging as a field developing and what might this type of computer science research offer for approaches to cancer diagnosis and treatment?
The application of deep learning methodologies in the field of medical imaging is, of course, a key aspect of modern predictive oncology. Through recent approaches based on the use of multi-layered convolutional neural networks and deep recurrent networks, it is now possible to extract, from medical images, features, and correlations previously unknown to medical experts. These approaches allow researchers and practitioners to better characterize the neoplasm both in the screening phase and in the follow-up to the pharmacological treatment.
Furthermore, by developing ad-hoc pipelines that include deep learning systems and advanced mathematical models, it is now possible to estimate the response to a given treatment either from diagnostic imaging (Radiomics) and/or from histology (Pathomics), thus allowing an optimal selection of the best drug treatment choice for the individual patient.
What kinds of studies or techniques are you looking for authors to contribute here?
We are interested in collecting studies, preliminary results, developments and pipeline proposals that exploit the potential of artificial intelligence and advanced mathematical modeling for the solution of issues in the field of medical oncology: from cancer early diagnosis, to predicting response to treatment, to oncological characterization of the neoplastic lesion to the assessment of the risk of relapse.
What impact do you think this collection will have in the field? How do you expect this collection will be used?
We hope and foresee an important impact in the scientific community considering that the present collection is addressed not only to researchers who are experts in the field of Computer Science and Artificial Intelligence, but also (and above all) to medical experts oncologists, radiologists and actually of any specialization which aim to contribute in the field of cancer control. We foresee the practical use of the pipelines collected in this collection through scientific sharing among researchers who submit their scientific contributions.
Why did you choose PeerJ to host this collection?
From an analysis of the papers already published in the PeerJ journals and considering the excellent editorial position of the same, we believe that PeerJ is the optimal journal to host this collection. We considered that the collection is aimed at an audience of scientific researchers of various disciplines, including engineers, mathematicians, doctors, biologists, etc.
When submitting to this collection, authors should state “Submitted to the Bio-inspired Deep Learning Image and Signal Processing Pipelines in Medical Oncology Collection” in the “Confidential Note to Staff” field in the submission form. General guidelines for PeerJ submissions can be found here. All submissions must meet PeerJ’s editorial criteria, policies and procedures, including data availability.