Success Story Topazium: AI for More Human-Centered Care By José Luis Diez, independent consultant and advisor in information technology, digital transformation, and digital health; business angel for startups. The company incorporates machine learning mechanisms to identify, for each patient’s profile, the combination of foods and ingredients that can best contribute to a healthy life. …
Advanced Prostate cancer – Docetaxel: Treatment with docetaxel improves survival in mCRPC but is associated with significant toxicity. The question remains as to whether the improved survival is worth the toxicity risk. Based on baseline clinical information, we have developed a simplified predictive tool to identify those patients on which treatment with docetaxel is likely to be beneficial.
Advanced colorectal cancer – Anti-EGFR antibodies: Practice guidelines recommend using anti-EGFR antibodies (e.g., cetuximab, panitumumab, etc.) to treat K-Ras wild-type mCRC patients where it was shown to significantly extend overall survival (OS). Still, a proportion of patients will not achieve this goal. Based on baseline clinical information, we have developed a simplified predictive tool to identify patients who are likely to benefit from anti-EGFR treatment.
LivGuard: is an AI-powered framework designed to identify signs of cirrhosis in liver ultrasound images. Engineered to be user-friendly, the system enables health professionals -regardless of their level of training- to upload images and receive results processed by an advanced deep neural network. It highlights regions likely to be cirrhotic, expediting detection and broadening access to medical imaging expertise, especially in regions with limited availability of hepatology specialists.
Signal Healer: is a powerful tool designed to address missing values in datasets. By performing supervised training on a tailored deep neural network, it uncovers variable dependencies across features, ensuring data integrity and reliability. The current version assumes independence between rows and treats all features as continuous. Future updates will expand its capabilities to handle diverse data types, including time series, images, and graphs, making it an even more versatile solution for AI developers.
HealthGuard: As chronic diseases increase with an aging population, Topazium’s HealthGuard steps in to transform nutrition and health. By pairing data from smartwatches and wearables with healthy foods, HealthGuard generates personalized food recommendations tailored to an individual’s health profile. This innovation not only enhances nutrition but also creates a pathway to improved quality of life and reduced risk of chronic diseases.
Patient-Drug Mapper: A significant number of hi-potential drugs end up falling through the cracks given their testing on patients with incompatible individual profiles. Topazium’s analytics allow the mapping of genetic and clinical patient’s conditions to a whole universe of experimental and approved drugs. Also, it can be used to produce a rank of matching alternative drugs given a patient’s profile, aiming both to increasing treatment efficiency whilst allowing drug re-profiling of vintage failed projects.
Pediatric Hand/Wrist X-rays: Traditional methods for calculating bone age using X-rays of the left wrist, hand, and fingers can be time-intensive and cumbersome. Topazium’s innovative solution streamlines this process with an algorithmic framework that provides instant age estimations. With a mean absolute error of approximately four months, this tool delivers rapid and reliable assessments of skeletal maturity, making it a valuable resource for pediatric healthcare professionals.
Radiology Chest X ray: Topazium’s advanced framework calculates an individual’s biological age based on chest X-ray data, identifying key factors that differentiate biological from chronological age. This tool empowers healthcare professionals to recommend targeted interventions, bridging the gap between these two ages and advancing personalized care.
Pneumothorax Diagnostician: Clinical suspicion of pneumothorax is usually confirmed by a radiologist based on a chest X-ray, a procedure which could be challenging if performed by professionals without extensive training. Topazium has implemented a framework that can detect and localize a pneumothorax in a given chest X-ray unsupervisely.
Thorax X-Rays Lung cancer: Topazium has developed a state-of-the-art deep learning network to detect and highlight signs of lung cancer on frontal chest radiographs. By generating heatmaps and providing probability assessments for suspicious regions, the system assists in the accurate identification of lung cancer presence or absence. This breakthrough technology supports healthcare professionals in early lung cancer detection -critical for improving outcomes for this severe disease.
IngrID (Ingredient Identifier): is a cutting-edge tool designed to match the most suitable food ingredients to achieve specific nutritional benefits. Whether the goal is to maintain a healthy lifestyle or prevent imbalances in physiological parameters, IngrID empowers nutraceutical researchers to uncover natural products rich in bioactive compounds. With IngrID, personalized nutrition takes a leap forward, helping people stay ahead in the pursuit of wellness.
Today, people are living longer than ever before. Global life expectancy at birth has risen to 73.3 years, with healthy life expectancy reaching 63.7 years. However, this increased longevity brings new challenges, as non-communicable diseases (NCDs) have become the leading health risks for aging populations. NCDs are responsible for approximately 43 million deaths annually, representing 75% of all global deaths. According to the WHO, early detection and treatment of NCDs are crucial for mitigating their impact, both in improving health outcomes and reducing economic burdens. The WHO’s strategy for controlling NCDs is holistic and multidisciplinary, focusing on four key stages:
Topazium aligns with the WHO’s vision by harnessing the power of AI technology to address these four stages. By integrating advanced AI with human expertise, Topazium develops innovative solutions to reduce the incidence, morbidity, and mortality of NCDs, ultimately improving quality of life on a global scale.
Data source: https://www.who.int/news-room/fact-sheets/detail/noncommunicable-diseases
Advanced eNSCLC Cancer Gemcitabine/Cisplatin combination: For individuals diagnosed with advanced squamous non-small cell lung cancer and exhibiting a favourable performance status, the initial therapeutic recommendation typically involves chemotherapy utilizing gemcitabine and cisplatin. However, an ongoing inquiry persists regarding the balance between the enhanced survival rates (showing a 20% decrease in the risk of mortality compared to optimal supportive care) and the potential for treatment-related toxicity. By leveraging deep computational models, Topazium’s tool has the capability to pinpoint patients who are predisposed to exceptional responses or prolonged survival following administration of this chemotherapy combination.
Biological Age Predictor: Topazium has designed and implemented a framework to accurately estimate a subject’s biological age from routine laboratory tests’ outputs. Additionally, it identifies the key metrics responsible for the potential basis between biological and chronological age, allowing professionals to prescribe rectifying actions aiming to close the referred biological/chronological age gap.
Dermatology SkinFace: This AI-powered application evaluates facial skin age using frontal photos, helping healthcare professionals align perceived age with biological skin health. The system generates heatmaps and probability scores to highlight areas of altered aging, supporting nutricosmetic approaches. By targeting specific nutrients and active ingredients, it promotes improved skin health and wellness tailored to individual needs.
Dermatology SkinHand: Topazium’s cutting-edge technology analyses hand images to accurately determine skin age, offering healthcare professionals a deeper understanding of the skin’s true biological condition. This innovation complements nutricosmetic strategies by leveraging nutrients and active ingredients in topical products to improve skin health and overall well-being, bridging the gap between appearance and biological reality.
Thorax X-Rays Global: Chest radiography is one of the most frequently performed radiological examinations worldwide. Powered by advanced deep neural networks, Topazium has developed an algorithmic framework capable of detecting and localizing abnormal radiological signs on frontal chest radiographs. The system also generates a heatmap, highlighting the regions of the image where abnormalities are detected. Designed to enhance access to medical imaging expertise, this tool is particularly valuable in regions with limited access to skilled radiologists, bridging the gap in healthcare delivery.
COVID-19 detector: Our tool uses a deep convolutional network to detect and locate radiological signs associated with COVID-19 in frontal chest radiographs, differentiating them from those associated with common bacterial and viral pneumonias. The physician supplies the anonymized radiological image via a cell phone, website, or email to the cloud. The system also generates a heatmap indicating the region of the image where the algorithm detects radiological signs associated with this pathology.
Lymph Node Metastasis Detector: Defining the presence or absence of lymph node metastases is essential in cancer diagnostics and staging, as it affects treatment and prognosis. This task is highly relevant but requires large amounts of reading time from a pathologist on-site or close to the point-of-care to analyze the samples. TOPAZIUM’s framework can detect cancer metastases in lymph nodes with an accuracy similar to the efficiency reported in the literature by expert pathologists.
Hereditary Angioedema: HADES is a predictive scoring system crafted to facilitate the early detection of hereditary angioedema (HAE), a rare condition frequently misdiagnosed due to overlapping symptoms with other disorders. By analyzing a select set of clinical variables, HADES offers a robust clinical performance. This simple yet powerful tool has the potential to transform HAE detection and improve patient outcomes.
Pigmented lesions: Skinguard is an advanced technological tool designed to assist healthcare professionals in improving their skills for detecting skin cancer. Simulating the teledermatology process, Skinguard analyzes a patient’s skin image and delivers a straightforward assessment by providing a “similarity score” that indicates how closely a lesion resembles either benign or malignant cases. It’s a valuable tool for enhancing precision in early skin cancer detection.
Diabetic retinopathy (DR) can affect anyone with hyperglycemia, though it primarily impacts people who have had diabetes for many years. It can cause permanent blindness if left undiagnosed and untreated. Clinicians can identify DR by ophthalmoscopy or retinopathy. While these approaches are effective, they require a high degree of expertise. Topazium has developed a framework that can automatically classify ocular fundus images according to the various severity degrees of DR.
Tumour Genetic Age: Cancer evolves differently based on the age of the patient, often leading to distinct tumour characteristics. Topazium’s Tumour Deep Clock tool analyzes the “mutational age” of tumours by profiling their gene signatures. By modeling cancer’s molecular landscape as a function of age, this tool identifies “young-tumour” and “old-tumour” genotypes, maps age-related biological changes, and connects these insights to clinical outcomes and tailored treatment approaches.
Microbiome Fingerprint: Topazium’s Microbiome Fingerprint tool harnesses machine learning to analyze vast microbiome datasets, identifying and categorizing microorganisms, predicting their roles, and exploring their interactions with the host. By translating intricate microbiome data into actionable clinical insights, this tool provides a powerful resource for advancing personalized medicine and understanding microbiome-driven health outcomes.
Methylation Fingerprint: Our Methylation Fingerprint tool deciphers complex DNA methylation patterns across the genome using cutting-edge machine learning algorithms. This framework links methylation to gene activity and phenotypic traits, helping researchers identify associations between methylation patterns and diseases or traits at the population level, offering new perspectives for precision medicine.
RNA Fingerprint: Topazium’s RNA Fingerprint tool applies advanced machine learning to RNA sequencing data, uncovering critical factors tied to specific health conditions. Using proprietary encoding methods, this system efficiently detects patterns and categorizes data, paving the way for the discovery of RNA biomarkers that drive disease progression and inform the development of targeted therapies.
Genetic Fingerprint: Our innovative machine learning framework leverages next-generation sequencing techniques -such as genome-wide, whole-exome, and gene-panel sequencing- to uncover data-driven insights into key genetic factors linked to specific health outcomes. With proprietary encryption methods, Topazium ensures secure genetic data analysis, enabling rapid pattern recognition, classification, and prediction. This approach identifies critical genetic biomarkers for the onset and progression of diseases like cancer, accelerating drug discovery, development, and personalized treatment strategies.
Virtual Target Identification: Identifying novel therapeutic targets is a critical step in successful drug discovery. To avoid the high costs associated with target validation, it’s essential to focus on candidates supported by strong scientific evidence. Topazium’s Virtual Target Identification tool harnesses vast genetic datasets to uncover proteins consistently linked to the onset of pathological conditions, streamlining the discovery process and driving precision in target selection.
Virtual Docking: When the physiological effects of a chemical entity are known, but the primary target remains unidentified, Topazium’s Virtual Docking steps in. This powerful tool uses cutting-edge neural network algorithms to explore interactions between candidates and specific targets, pinpointing the most likely match. Additionally, it reveals critical binding interactions within the target’s cleft, providing insights for chemical optimization and advancing drug development efforts.
Virtual Screening: The discovery of potential therapeutic molecules often requires testing an overwhelming number of compounds—a process that can be both time-consuming and costly. Topazium’s Virtual Screening tool leverages advanced deep learning strategies to identify compounds with the highest likelihood of interacting with a specific therapeutic target. By narrowing the focus to the most promising candidates, this tool ensures experimental efforts are invested where success is most probable, saving time and resources.
TNDR (Topazium Network for Drug Repurposing): redefines therapeutic innovation by connecting chemical compounds to diseases through advanced neural networks. Harnessing molecular disease data and chemical compound properties, TNDR discovers new therapeutic uses for existing drugs, identifies novel indications for emerging compounds, and highlights approved drugs to tackle unmet medical needs. Each compound-disease pair is assigned a match score, signaling its therapeutic potential and unlocking new horizons in drug repurposing.
GeneSight: uncovers the molecular pathways driving the physiological effects of chemical compounds. By providing in-depth molecular insights, GeneSight reveals the molecular mechanisms behind these effects, paving the way for groundbreaking discoveries in precision nutrition and pharmacology.
MetaPharm: Optimal nutrition is the cornerstone of health and well-being, forming the very foundation of metapharmacology. By leveraging state-of-the-art neural networks, MetaPharm analyzes vast global pharmacological data to identify food ingredients and components that can deliver targeted nutritional benefits. With MetaPharm, we aim to make precision nutrition a reality.
Melanoma: Diagnosis in dermatology is largely based on visual inspection of a lesion or the suspicious skin area, depending largely on the experience and training of the dermatologist (or, general practitioners in areas where dermatological services are not readily available). Topazium’s framework currently classifies seven common pigmented-skin lesions with an accuracy similar to the efficiency reported in the literature by well-trained dermatologists.
Who is the Data Controller?
The controller of your personal data is TOPAZIUM with registered office in Paseo de la Castellana 40, 8th floor, 28046, Madrid, Spain.
For any queries or requests you have about data protection, you can contact TOPAZIUM at the following email address: info@topazium.com
What are the purposes of processing personal data?
In the event that you have filled in your details in the “Contact Us” section, TOPAZIUM will process your personal data in order to respond to and manage your contact request by resolving the request for information you may have required.
TOPAZIUM will only process the personal data that you provide together with the message sent through the form.
The processing of this personal data by TOPAZIUM to resolve your query is based on the interest of TOPAZIUM to offer the best possible service to any third party interested in the activity of TOPAZIUM.
The TOPAZIUM website installs cookies on users’ browsing devices in order to ensure the proper functioning of the website and to analyze the behavior of users. The installation of cookies is carried out voluntarily by Users who can withdraw their consent at any time by configuring cookies through the control panel provided by TOPAZIUM. You can find more information about cookies in our Cookies Policy.
How long will we keep your data?
We will process your data for as long as necessary to respond to and resolve the request you have made to us through the Contact Us section.
In the case of cookies, the data may be processed as long as you maintain your consent for its installation and for the duration of the installation of the same.
To whom do we communicate your data?
TOPAZIUM has collaborators who to provide their services can access the personal data of TOPAZIUM.
TOPAZIUM makes a careful selection and control of these third parties with which it has signed data protection agreements, where they are obliged to comply with a number of data protection obligations.
TOPAZIUM may also communicate data of interested parties to the administrative authorities that require it.
What are the rights of your personal data and how to exercise it?
You may, in the terms established in the data protection regulations, revoke at any time the authorization granted for the processing and transfer of your personal data, as well as exercise your rights of access, rectification, opposition, limitation, deletion, portability and not be subject to automated decisions, writing to TOPAZIUM to the email address info@topazium.com or to the address Paseo de la Castellana 40, 8th floor, 28046 Madrid, Spain.
How can you enforce your data protection rights?
If you believe that any of your data protection rights have been violated, you may contact TOPAZIUM to rectify or end your concern about the matter at the address mentioned in the previous section. In any case, you can always contact the data protection authorities to safeguard your rights www.aepd.es