The data scientist as a mainstay of the tumor board: global implications and opportunities for the global south

Abstract

Tumor boards are multidisciplinary teams of healthcare professionals that are working together to encompass the full spectrum of care around diagnosing, planning treatment, and advising outcomes for individual cancer patients. These boards typically consist of oncologists, radiologists, pathologists, geneticists, surgeons, nurse practitioners, and other palliative care professionals (National Cancer Institute, 2024). These boards create a collaborative space for experts from various disciplines to assess clinical factors and patient circumstances, ensuring the application of appropriate care standards and personalized recommendations from the National Comprehensive Cancer Network (NCCN) Guidelines to enhance cancer treatment are met. Since no patient fits the "textbook" cancer profile, oncologists benefit from discussing tailored treatment plans and learning from their colleagues' experiences. When tumor boards are functioning well, they can have a significant impact on patient care (NCCN, 2025). For instance, a thoracic oncology board in Munich, Germany, found that 90% of their recommendations met or exceeded clinical standards, with nearly 90% being implemented in practice (Walter et al, 2023).Tumor boards are increasingly used worldwide, but expertise and resources for conducting multidisciplinary tumor boards are still limited in the Global South. However, this does not mean they cannot be implemented in developing countries. A 2020 survey from Southeast Asia found that 80.4% of pediatric solid tumor units had pediatric-trained specialists, including oncologists, surgeons, radiologists, pathologists, radiation oncologists, nuclear medicine physicians, and nurses. This indicates that multidisciplinary tumor boards are already in place and that these specialists play a critical role in cancer care (Ottman, 2020). With full implementation in the global south, data scientists can further enhance tumor boards with AI and data analytics to improve decision-making and personalize cancer care.Advances in big data, machine learning (ML), and artificial intelligence (AI) provide more precise, evidence-based, and patient-specific care, thus, giving a different approach as to how healthcare professionals diagnose, treat, and manage their patients (Alowais et al, 2023). For instance, there is a growing number and complexity of data in the healthcare industry such as from Electronic Health Records (EHRs), next-generation genomic sequencing (NGS), and advanced imaging modalities like X-ray Radiography, Magnetic Resonance Imaging (MRI), and Computed Tomography (CT) scans. However, analyzing these data, individually and manually, can be time-consuming and considerably impractical. This is where clinical decision support systems (CDSS) powered by AI and ML are put into action. These systems provide predictive analysis of the disease progression or prognosis, personalized treatment based on the patients' genomic profile, and drug-drug interaction alerts (Wang et al, 2023;Alowais et al, 2023). As precision medicine and big data continue to evolve, healthcare will increasingly rely on data-driven tools to enhance patient care, reduce errors, and improve overall health outcomes (Khalifa and Albadawy, 2024). Data scientists are critical to this process as they can analyze large datasets to identify biomarkers that can predict how a patient will respond to specific treatments (Nardone et al, 2024). In addition, AI algorithms are being used to interpret radiological images, detect early signs of cancer, and predict tumor progression. These tools are increasingly becoming standard in tumor boards, especially in high-income countries (Bi et al, 2019;El Saghir et al, 2015).For instance, in oncology, the most commonly used diagnostic tools to identify biomarkers and guide targeted therapies in precision medicine are Polymerase Chain Reaction (PCR), fluorescent in situ hybridization (FISH), and immunohistochemistry (IHC) to identify biomarkers and guide targeted therapies (Goosens et al, 2015). However, high-throughput next-generation sequencing (NGS)-based diagnostics, which analyze somatic mutations in tumors, have proven clinically useful in identifying single-nucleotide mutations, insertions, deletions, and large genomic rearrangements (Kamps et al, 2017). Thus, multigene NGS testing can provide the oncologist a clinical picture of the patients' molecular profile which can be utilized in planning the best treatment option (Mehta et al, 2020).As precision medicine continues to gain prominence and the molecular characterization of individual cancers becomes increasingly complex (Specchia et al., 2020;Nardone et al., 2024), incorporating data scientists into tumor boards is essential. Data scientists bring advanced expertise in ML, data analysis, and bioinformatics, enabling tumor boards to make more accurate, evidence-based clinical decisions that lead to improved patient outcomes (Nardone et al., 2024;Rodriguez Ruiz et al., 2022). They play a critical role in synthesizing and analyzing diverse datasets generated in oncology care, uncovering actionable insights, and informing treatment strategies. This role is particularly as cancer treatment personalized based on the and molecular of data scientists advanced including analysis, and predictive to actionable and treatment of such as from and to from data, such as and (Nardone et al, et al, there is a growing data-driven into cancer In high-income tools are being used to in images, patient and identifying treatment strategies. In the for like Cancer are AI tools to guide clinical decisions in oncology (Wang et al, 2023). These powered by machine learning can analyze patient data in and provide recommendations tumor board However, the of these tools with countries significant to to a of and expertise et al, high-income countries like in the and data scientists are already of tumor For instance, data scientists the of with oncologists to personalized treatment plans based on genomic data and machine learning 2024). These experts tools to predict patient to various therapies and identify clinical in Germany, data scientists with oncologists to analyze data and into clinical ensuring that patient the most appropriate care based on their profile et al, by data scientists have personalized cancer analyzing genomic and clinical data scientists can identify mutations cancer and targeted therapies that are to be more standard treatments and can be used to patient oncologists to treatment plans based on the These data-driven reduce the of cancer treatment and lead to more and care (National Cancer Institute, is a in with from and to and data the of these AI is to have a significant impact on the of et al, in to the of and to healthcare especially This of to their to AI in or other healthcare et al, the most significant and in AI are not in and advanced but in and with the healthcare and high-income countries have significant in data into oncology care, significant These a of to and a of professionals of AI and machine learning tools et al, 2020). there are and the of AI in particularly in countries with diverse patient and 2022). these will significant in and as as the of for the of AI in clinical the Global which or developing in and or the cancer care is by a of to advanced diagnostic a of healthcare and These to of cancer and outcomes to high-income countries and 2024). In of and Southeast patients to to in and treatment et al, 2023). healthcare systems in these are and the resources to the in cancer like in the the of cancer care is the of to global standards of care to the to the patients and their 2023). global and recommendations to precision medicine in the care of patients with cancer in the et al, is still for to this in their clinical not a oncologists in the AI in precision medicine in their patients et al, scientists have the to healthcare in the Global by AI and machine learning to and improve diagnostic such as and health can be used to provide diagnostic support in where to healthcare professionals is limited et al, et al, 2024). ML can be used to predict cancer identify patients of developing and most in of treatment 2023). for AI for oncology et al, is found to be a in cancer care in et al, 2024). these tools can disease progression and treatment enabling more personalized care (Bi et al, developing predictive that can be in the of advanced diagnostic data scientists can improve outcomes for cancer patients in these there are discussing the of AI in healthcare in the full implementation of AI in healthcare or in oncology are In for instance, are being used to for cancer in the to and in machine learning algorithms have used to analyze to with et al, 2024). These the for data to cancer care in the Global by to of the most the already to the by identifying the to AI implementation in and the that can most the appropriate of AI to enhance health outcomes in these as a critical approach in particularly in cancer treatment, where of expertise are to complex from various This collaborative approach professionals from diverse such as bioinformatics, data and from these healthcare can more and personalized care for especially in oncology 2017). For instance, in the of in and other healthcare professionals such as oncologists, and have together to identify the of patients and to et al, 2017). teams have proven to be a in patient the of and other health the to patients' more ensuring of care et al, a analysis of the on that with multidisciplinary teams outcomes 2017). are increasingly found including to data such as a health in Thus, these the for data scientists to make a significant impact teams various healthcare to tumor boards, data scientists a of and A solid in data analysis, and machine learning is particularly with expertise in datasets such as data, and health a in a with a on and data such as or This of expertise the to large and datasets to healthcare to the Health and of the to patient data and of and clinical with oncologists, radiologists, pathologists, and other healthcare data scientists in complex into actionable insights, data that diverse and are for into the multidisciplinary of tumor the and of their for this role by or a to learning in data and oncology that data scientists can to and in precision from the of and on the of as a particularly in in the Global the critical of and in expertise into This that as and have advanced healthcare limited resources to these to the of data scientists in tumor boards the for and or tailored to to the and of data scientists such as the Data which a for the of data such as the the of the and the et al., et al., et al., et al., In the the for health which with data et al., for further this with for data and for the for AI and data with a on of 2024). the National of the a standard which for data analysis in particularly to data (National data a role as a of from datasets including and data et al, 2022). tumor boards, data scientists not advanced ML but a of and to treatment planning and personalize patient care (Nardone et al, 2024). of their predictive and data into more such as and a of machine learning where to make of decisions by growing in treatment planning et al, 2020). Data scientists to for the the oncology and a the that the of a on et al, They this by to a In oncology, is used to treatment that based on the to algorithms can of radiation or to et al, a based on for personalized proven in the of with AI in early and can be to different et al, 2022). A personalized treatment approach to treatment plans in as patient data personalized to individual early and treatment scientists that a of and their These in outcomes and the of various treatment based on In a tumor board can from different data, molecular and patient the of different clinical outcomes and provide make more decisions in where there is et al, 2024). For in the of et they a patient to patient data and for treatment decisions in and the with and significant to treatment decisions for patients. This to a decision-making in tumor boards, clinical and molecular data to identify the best treatment for data scientists to create treatment and can the of various plans to patients. These are especially in complex with of treatments like with or predict patient based on 2022). like in the of et they a for Tumor treatment planning for on the tumor with their based on patient and tumor such algorithms into advanced treatment planning systems is to enhance of patient to treatment data scientists to personalize the and of therapies such as or These to with especially in treatments that based on individual patient and tumor can and to the of patient of 2024). of treatments further for strategies. Data scientists can be useful by which can approach to assess and a clinical et al, 2023). in learning have to the of from patient or to treatment planning systems for can predict based on data and et al, is of treatment Data scientists data from and to create a molecular profile of a these diverse data scientists tumor boards more targeted For a approach not a the cancer but how with other molecular to more tailored such as in to the genomic and 2023). a of cancer but the large data AI these by identifying and from complex data, cancer et al, 2024). to the by et which machine learning for data in cancer their on and They data from the Cancer in cancer and recommendations for appropriate and the of tools to and personalized from imaging that can tumor and predict patient to identify that not be by a Data scientists these with clinical and genomic data to improve treatment and strategies. in the of et they ML that can and to predict in patients with and as as their overall (Wang et al, et al, and are particularly in where testing not be for tumor et al, et al, 2022). such as and are cancer by molecular with These the of tumor and that sequencing cannot For instance, like and enhance tumor analysis by imaging with and et al., 2023). enabling more of and in et al., of the most of in the of oncology is the of in the improved patient outcomes in precision cancer treatment still to where different of the respond to that the from the to the that tumor can respond on their et al, 2024). Thus, is to not the profile of the tumor but and the This approach lead to more personalized and treatments by the and the of treatment on to how specific lead to This is particularly critical in oncology, where data scientists advanced like the et al., and to identify in treatment and disease progression et al., 2022). like the and interpret machine learning they et al., a large used to improve of the factors in Cancer such as a on treatment et al, 2024). role of data scientists such as and to like treatment in analysis et al., these into clinical data scientists tumor boards to decisions on insights, patient outcomes in personalized oncology application where data scientists play a role is in developing treatment strategies. These are to in to the For instance, a tumor becomes to data scientists can to such as to targeted or analyzing patient data the treatment that treatment plans and to in disease progression et al, For is used as a cancer treatment that the radiation treatment to for in the and as as This make the treatment more and by the tumor to the and et al, scientists decision support systems that oncologists in treatment based on the of patient data, clinical and outcomes from These systems a of machine learning algorithms and clinical to provide recommendations that standard care and personalized In data scientists systems of decision support by data from or For a clinical the in the to or et al, data in cancer patients for such as and in and treatment identify and et al, 2024). a from the of found that especially be used to create a clinical (CDSS) for cancer decision support 2024). A of or radiation the of and a mean of the tumor decision as scientists can where treatments such as and are used in the interaction various they can predict the or outcomes of This is particularly in oncology, where treatments are but be to benefit et al, 2024). For instance, are used to improve treatment precision and reduce by especially in with and other cancer treatment et al, 2025). With the of data predictive can be utilized to in the and of and This will enhance the of their in and cancer decisions for and cancer by patient data and treatment et al, clinical and data in Network to et al, with advanced to Network and enhance predictive et al, Tumor for by and tumor et al, 2020). particularly like and et al, et al, et al, et al, have and in tools for tumor These are clinical data, including and images, which critical for and analyzing of data, to make more precise, data-driven decisions that improve patient outcomes et al, including and et al, et al, large of data from patient and clinical to data and complex like and analysis significant which with the and in data 2024). In tumor boards, can complex patient and predict treatment oncologists with actionable are to enhance for the of et al, This is particularly for the progression of like cancer, where data tumor and treatment can care strategies. are in treatment assess how a et al, play a role in factors such as and et al, These analyze complex to provide into molecular that cancer progression. identify for personalized treatment in oncology et al, a approach to et al, These are how a useful for treatment and disease progression. are particularly in like oncology, where patient can to treatment or disease is a in to in the in et al, 2023). In tumor boards, can be to analyze patient such as data from clinical or This to for from patient and radiological to more personalized and treatment the found in with which are and are particularly useful in where and data is such as analyzing large datasets of patient or data from et al, 2022). In tumor boards, be used to from data like patient in the for more decision-making cancer are to process data large like but with more et al, 2023). This can be useful in oncology, where a of patient data is the can tumor boards the of treatments analyzing data from and genomic to that treatments as the disease is a in that and 2017). is particularly in where different of a of In tumor boards, these can resources to the most critical patient data, such as genomic or in patients with advanced This that the on the most data to with clinical play role in data, which is in oncology to the complexity of and clinical and can be used to these large datasets the most for tumor or In tumor boards, by data enabling to on the most critical for personalized treatment planning et al, on which to analyze complex and for in the complexity of cancer data, especially as datasets in and complexity et al, 2023). learning particularly a more as they can from data and in These are not by the of and are to the of cancer treatment et al, working data scientists advanced ML the and data-driven This that tumor boards can the most insights, for more and personalized cancer care around that the of healthcare teams to more diverse as data AI specialists, and to improved patient As healthcare becomes more the of data scientists are to further et al, like the learning health that learning and data lead to clinical which with the data role in synthesizing diverse of et al, professionals are to and AI in their to the and of As of the healthcare for instance, and in AI to for the of This can be most AI and are into the ensuring are with the and tools to and AI in their practice et al, 2024). AI can support by their to provide and imaging for with other healthcare and decision AI on more complex of patient care and and teams that are to complex as they benefit from the expertise of professionals from different This collaborative approach the to to and patient of is and most are to on these in to their Data scientists are with the as and from and datasets to the tumor This tumor boards to the of cancer treatment such as in the Global a data to to from limited data can be these to data, data scientists can improve diagnostic treatment and make personalized cancer care more to et al, Global significant to the of data in particularly in is a of in the Global by is by where indicates the to For the can be to the most in the such as in the with based on the in there is not a AI which can from For the which number on the be the for AI which be used for AI be used for predictive of such as of of the the for AI in tumor of is that does not for systems to is to the of which is the by with or This of the on the and for such as is developing AI but the of to the a of with is in developing this but countries which are and These countries can the to the to a and AI Southeast Asia is of AI including Asia including and and with the is the in developing in the especially with the of the with of and a of in other countries in the are for the of such including and the will be to this is a of data scientists and healthcare professionals in advanced data analytics in healthcare the to complex AI the of data in clinical practice et al, in Health Data Health systems in from with limited in health This of support the of health data systems that the of AI and ML in cancer care and of AI and have as to AI to improve healthcare in of the to AI have with and such as to 2024). and AI the to improve healthcare in and is as a for global health et al, et al, 2024), data in the to and For the data and developing countries to a in which of data from or and to which are to AI tools be used to further a such as the and and a the to this and and and with to with the of AI is that on to This can in outcomes like and medicine et al, These AI being can still make or in a the is for ensuring the best the other patients AI and most of are with are on AI for This is by factors such as and on For in or patients are AI they that will their or no significant They are in AI for health support a they that AI does not However, with and more in AI AI as a for patient Thus, have from limited and data addition, patients are more their health have and is on a and on and This support patients' and from Thus, AI with these et al, in they a a to cancer and a to the and are precise, by that the the is AI with most oncologists not being by be that and healthcare professionals rely on AI in the of or Thus, data scientists be of the role they play in As data scientists more in tumor boards, they how AI can enhance the patient by ensuring they and Data scientists and have of data scientists on such as data analysis, AI and patient care, and clinical These can lead to in working as they not or data scientists and to Data scientists to complex not the of patient care, to and in data-driven healthcare these to and together more Data scientists can their to make their more to can data scientists the of patient This that AI and data tools are in that benefit patients 2022). of For oncologists and data a is data as a to enhance patient care, data scientists gain a of clinical and are with data scientists their and oncologists data-driven from This can lead to more decision-making and improved patient outcomes by clinical with data-driven learning is a global that data and with and datasets et al, et al, et al, et al, 2022). data from various but in of the overall data being to other of implementation et al, and et al, 2022). As and in developing countries the for data and with as a to this the data and a in the of a in which for as they with to their AI play a role in the by where to advanced These to on or with limited have significant various For instance, et for enabling and more in healthcare a AI tailored to reduce health in and like a how can disease in critical in healthcare et to healthcare by and AI to and diagnostic healthcare more and for the of AI that not but with the of As et these into and to like data and the for AI in further how AI can in For instance, machine learning and data enabling AI to be more and et al., 2024). as by et and for machine learning in have in on AI for decision-making et al., et al., 2025). not enhance but in healthcare ensuring that benefit from These the of AI to improve of support and in the of data in cancer care, from are play role in in health data systems and can create enabling by that ensuring data and and for health Health can as healthcare systems and for in health can to by and et al, especially in the and healthcare can provide expertise and for the of AI and data in These can such as data and AI that are more in the the Global a to of cancer care by AI and data can the for by and health These healthcare to and analyze data cancer care more and in et al, 2024). Data and data health and can datasets that into cancer care These can be used to treatment that the and disease in these et al, et al, 2024). and diagnostic tools can improve early by analyzing imaging and data with AI can personalize treatment plans based on disease and enabling more targeted cancer therapies et al, data and artificial intelligence (AI) in especially in for and these can healthcare professionals enhance their that can be can guide this healthcare professionals as they and as by a to approach with and In this the to and how data-driven or enhance This that by on their and professionals can make AI and data learning can healthcare teams these as useful tools in patient care and especially tumor boards where data-driven decisions can have a However, the of and that professionals are growing and In this the of to that data and AI and can the by and this as to on by that can further enhance patient the of as for In professionals approach AI and data with a are to these tools as a to their and improve their to care for patients. a teams a where professionals in learning the of AI and data a a and data scientists into tumor boards in and this and to these a and for healthcare systems with the expertise to data-driven cancer care with a of the specific by healthcare and be with tumor boards, healthcare and to identify in and For to clinical and for a of including healthcare and the is with patient further with healthcare the for and on global standards like the Data et al., et al., et al., et al., this oncology, data and machine tumor predictive and in data the and of ensuring with healthcare and such as or can the of tailored to is in data scientists with and can the of and learning like or tailored for can provide learning be by to tumor board tools and healthcare such as tumor in as a testing for These be in with

Ähnliche Arbeiten