Carlos Aumente-Maestro
PhD Student
Carlos is a staff scientist at Helmholtz Imaging and the Intelligent Medical Systems (IMSY) division at the German Cancer Research Center (DKFZ). His work focuses on developing open-source tools and standardized evaluation methodologies to enable robust and reproducible validation of medical image analysis systems. His research interests lie at the intersection of trustworthy AI, validation methodology, and the clinical translation of medical imaging algorithms.
Before joining Helmholtz Imaging, he worked for several years in industry as a data scientist and machine learning engineer. He is currently completing a PhD focused on improving biomedical image analysis through deep learning techniques. As part of his PhD, he has explored topics such as lightweight models for brain tumor segmentation, multitask systems for breast cancer segmentation and classification, and has developed AUDIT (Analysis and evalUation Dashboard of artIficial inTelligence), an open-source Python library for comprehensive performance assessment and MRI dataset characterization in medical image segmentation.
Publications
4725570
Aumente-Maestro
1
https://helmholtz-imaging.de/apa-bold-title.csl
50
date
desc
5335
https://helmholtz-imaging.de/wp-content/plugins/zotpress/
%7B%22status%22%3A%22success%22%2C%22updateneeded%22%3Afalse%2C%22instance%22%3Afalse%2C%22meta%22%3A%7B%22request_last%22%3A0%2C%22request_next%22%3A0%2C%22used_cache%22%3Atrue%7D%2C%22data%22%3A%5B%7B%22key%22%3A%222G4WKZR2%22%2C%22library%22%3A%7B%22id%22%3A4725570%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Aumente-Maestro%20et%20al.%22%2C%22parsedDate%22%3A%222025-08-05%22%2C%22numChildren%22%3A1%7D%2C%22bib%22%3A%22%26lt%3Bdiv%20class%3D%26quot%3Bcsl-bib-body%26quot%3B%20style%3D%26quot%3Bline-height%3A%201.35%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%26quot%3B%26gt%3B%5Cn%20%20%26lt%3Bdiv%20class%3D%26quot%3Bcsl-entry%26quot%3B%26gt%3BAumente-Maestro%2C%20C.%2C%20M%26%23xFC%3Bller%2C%20M.%2C%20Remeseiro%2C%20B.%2C%20D%26%23xED%3Bez%2C%20J.%2C%20%26amp%3B%20Reyes%2C%20M.%20%282025%29.%20%26lt%3Bb%26gt%3BAUDIT%3A%20An%20open-source%20Python%20library%20for%20AI%20model%20evaluation%20with%20use%20cases%20in%20MRI%20brain%20tumor%20segmentation%26lt%3B%5C%2Fb%26gt%3B.%20%26lt%3Bi%26gt%3BComputer%20Methods%20and%20Programs%20in%20Biomedicine%26lt%3B%5C%2Fi%26gt%3B%2C%20108991.%20%26lt%3Ba%20class%3D%26%23039%3Bzp-DOIURL%26%23039%3B%20href%3D%26%23039%3Bhttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1016%5C%2Fj.cmpb.2025.108991%26%23039%3B%26gt%3Bhttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1016%5C%2Fj.cmpb.2025.108991%26lt%3B%5C%2Fa%26gt%3B%26lt%3B%5C%2Fdiv%26gt%3B%5Cn%26lt%3B%5C%2Fdiv%26gt%3B%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22AUDIT%3A%20An%20open-source%20Python%20library%20for%20AI%20model%20evaluation%20with%20use%20cases%20in%20MRI%20brain%20tumor%20segmentation%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Carlos%22%2C%22lastName%22%3A%22Aumente-Maestro%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Michael%22%2C%22lastName%22%3A%22M%5Cu00fcller%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Beatriz%22%2C%22lastName%22%3A%22Remeseiro%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Jorge%22%2C%22lastName%22%3A%22D%5Cu00edez%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Mauricio%22%2C%22lastName%22%3A%22Reyes%22%7D%5D%2C%22abstractNote%22%3A%22Background%3A%5CnThe%20development%20of%20artificial%20intelligence%20%28AI%29%20models%20for%20medical%20image%20analysis%20remains%20a%20significant%20challenge.%20Main%20issues%20include%20a%20lack%20of%20model%20generalization%2C%20the%20need%20for%20extensive%20datasets%20for%20training%2C%20and%20poor%20model%20robustness%20for%20clinical%20application.%20Current%20evaluation%20frameworks%20often%20fail%20to%20provide%20subject-based%20insights%20necessary%20to%20identify%20model%20biases%20and%20address%20domain%20shifts.%5CnMethods%3A%5CnTo%20address%20these%20limitations%2C%20this%20study%20introduces%20AUDIT%2C%20an%20open-source%20Python%20library%20designed%20to%20enhance%20the%20evaluation%20of%20segmentation%20models%20and%20the%20analysis%20of%20MRI%20datasets.%20AUDIT%20includes%20modules%20for%20extracting%20region-specific%20features%20and%20calculating%20performance%20metrics%2C%20as%20well%20as%20a%20dynamic%20web%20APP%20for%20interactive%20model%20evaluation%20and%20data%20exploration.%5CnResults%3A%5CnAll%20the%20source%20code%2C%20tutorials%2C%20and%20documentation%20can%20be%20found%20at%20https%3A%5C%2F%5C%2Fgithub.com%5C%2Fcaumente%5C%2FAUDIT%2C%20and%20can%20directly%20be%20installed%20from%20the%20Python%20Package%20Index%20using%20pip%20install%20auditapp.%20Through%20a%20series%20of%20use%20cases%20commonly%20encountered%20in%20AI-driven%20brain%20tumor%20segmentation%2C%20this%20manuscript%20highlights%20the%20versatility%20and%20broad%20applicability%20of%20the%20AUDIT%20library%20across%20diverse%20scenarios.%5CnConclusions%3A%5CnAUDIT%20was%20designed%20to%20fill%20the%20existing%20gaps%20in%20the%20literature%20on%20the%20evaluation%20of%20AI%20segmentation%20models%2C%20ultimately%20aiming%20to%20advance%20the%20AI%20medical%20image%20analysis%20domain.%20Beyond%20its%20extensive%20analysis%20capabilities%2C%20AUDIT%20has%20been%20designed%20to%20support%20integration%20with%20external%20libraries%20and%20applications.%22%2C%22date%22%3A%222025-08-05%22%2C%22language%22%3A%22%22%2C%22DOI%22%3A%2210.1016%5C%2Fj.cmpb.2025.108991%22%2C%22ISSN%22%3A%220169-2607%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fwww.sciencedirect.com%5C%2Fscience%5C%2Farticle%5C%2Fpii%5C%2FS0169260725004080%22%2C%22collections%22%3A%5B%5D%2C%22dateModified%22%3A%222025-08-07T08%3A54%3A02Z%22%7D%7D%2C%7B%22key%22%3A%223YQG6PQE%22%2C%22library%22%3A%7B%22id%22%3A4725570%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Aumente-Maestro%20et%20al.%22%2C%22parsedDate%22%3A%222025-03-01%22%2C%22numChildren%22%3A1%7D%2C%22bib%22%3A%22%26lt%3Bdiv%20class%3D%26quot%3Bcsl-bib-body%26quot%3B%20style%3D%26quot%3Bline-height%3A%201.35%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%26quot%3B%26gt%3B%5Cn%20%20%26lt%3Bdiv%20class%3D%26quot%3Bcsl-entry%26quot%3B%26gt%3BAumente-Maestro%2C%20C.%2C%20D%26%23xED%3Bez%2C%20J.%2C%20%26amp%3B%20Remeseiro%2C%20B.%20%282025%29.%20%26lt%3Bb%26gt%3BA%20multi-task%20framework%20for%20breast%20cancer%20segmentation%20and%20classification%20in%20ultrasound%20imaging%26lt%3B%5C%2Fb%26gt%3B.%20%26lt%3Bi%26gt%3BComputer%20Methods%20and%20Programs%20in%20Biomedicine%26lt%3B%5C%2Fi%26gt%3B%2C%20%26lt%3Bi%26gt%3B260%26lt%3B%5C%2Fi%26gt%3B%2C%20108540.%20%26lt%3Ba%20class%3D%26%23039%3Bzp-DOIURL%26%23039%3B%20href%3D%26%23039%3Bhttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1016%5C%2Fj.cmpb.2024.108540%26%23039%3B%26gt%3Bhttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1016%5C%2Fj.cmpb.2024.108540%26lt%3B%5C%2Fa%26gt%3B%26lt%3B%5C%2Fdiv%26gt%3B%5Cn%26lt%3B%5C%2Fdiv%26gt%3B%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22A%20multi-task%20framework%20for%20breast%20cancer%20segmentation%20and%20classification%20in%20ultrasound%20imaging%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Carlos%22%2C%22lastName%22%3A%22Aumente-Maestro%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Jorge%22%2C%22lastName%22%3A%22D%5Cu00edez%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Beatriz%22%2C%22lastName%22%3A%22Remeseiro%22%7D%5D%2C%22abstractNote%22%3A%22Background%3A%5CnUltrasound%20%28US%29%20is%20a%20medical%20imaging%20modality%20that%20plays%20a%20crucial%20role%20in%20the%20early%20detection%20of%20breast%20cancer.%20The%20emergence%20of%20numerous%20deep%20learning%20systems%20has%20offered%20promising%20avenues%20for%20the%20segmentation%20and%20classification%20of%20breast%20cancer%20tumors%20in%20US%20images.%20However%2C%20challenges%20such%20as%20the%20absence%20of%20data%20standardization%2C%20the%20exclusion%20of%20non-tumor%20images%20during%20training%2C%20and%20the%20narrow%20view%20of%20single-task%20methodologies%20have%20hindered%20the%20practical%20applicability%20of%20these%20systems%2C%20often%20resulting%20in%20biased%20outcomes.%20This%20study%20aims%20to%20explore%20the%20potential%20of%20multi-task%20systems%20in%20enhancing%20the%20detection%20of%20breast%20cancer%20lesions.%5CnMethods%3A%5CnTo%20address%20these%20limitations%2C%20our%20research%20introduces%20an%20end-to-end%20multi-task%20framework%20designed%20to%20leverage%20the%20inherent%20correlations%20between%20breast%20cancer%20lesion%20classification%20and%20segmentation%20tasks.%20Additionally%2C%20a%20comprehensive%20analysis%20of%20a%20widely%20utilized%20public%20breast%20cancer%20ultrasound%20dataset%20named%20BUSI%20was%20carried%20out%2C%20identifying%20its%20irregularities%20and%20devising%20an%20algorithm%20tailored%20for%20detecting%20duplicated%20images%20in%20it.%5CnResults%3A%5CnExperiments%20are%20conducted%20utilizing%20the%20curated%20dataset%20to%20minimize%20potential%20biases%20in%20outcomes.%20Our%20multi-task%20framework%20exhibits%20superior%20performance%20in%20breast%20cancer%20respecting%20single-task%20approaches%2C%20achieving%20improvements%20close%20to%2015%25%20in%20segmentation%20and%20classification.%20Moreover%2C%20a%20comparative%20analysis%20against%20the%20state-of-the-art%20reveals%20statistically%20significant%20enhancements%20across%20both%20tasks.%5CnConclusion%3A%5CnThe%20experimental%20findings%20underscore%20the%20efficacy%20of%20multi-task%20techniques%2C%20showcasing%20better%20generalization%20capabilities%20when%20considering%20all%20image%20types%3A%20benign%2C%20malignant%2C%20and%20non-tumor%20images.%20Consequently%2C%20our%20methodology%20represents%20an%20advance%20towards%20more%20general%20architectures%20with%20real%20clinical%20applications%20in%20the%20breast%20cancer%20field.%22%2C%22date%22%3A%222025-03-01%22%2C%22language%22%3A%22%22%2C%22DOI%22%3A%2210.1016%5C%2Fj.cmpb.2024.108540%22%2C%22ISSN%22%3A%220169-2607%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fwww.sciencedirect.com%5C%2Fscience%5C%2Farticle%5C%2Fpii%5C%2FS0169260724005339%22%2C%22collections%22%3A%5B%5D%2C%22dateModified%22%3A%222025-08-07T08%3A55%3A36Z%22%7D%7D%2C%7B%22key%22%3A%22P2QAW5FX%22%2C%22library%22%3A%7B%22id%22%3A4725570%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Aumente-Maestro%20et%20al.%22%2C%22parsedDate%22%3A%222025%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%26lt%3Bdiv%20class%3D%26quot%3Bcsl-bib-body%26quot%3B%20style%3D%26quot%3Bline-height%3A%201.35%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%26quot%3B%26gt%3B%5Cn%20%20%26lt%3Bdiv%20class%3D%26quot%3Bcsl-entry%26quot%3B%26gt%3BAumente-Maestro%2C%20C.%2C%20Gonz%26%23xE1%3Blez%2C%20D.%20R.%2C%20Martinez%2C%20D.%2C%20%26amp%3B%20Remeseiro%2C%20B.%20%282025%29.%20%26lt%3Bi%26gt%3B%26lt%3Bb%26gt%3B%26lt%3Bspan%20style%3D%26quot%3Bfont-style%3Anormal%3B%26quot%3B%26gt%3BBTS%20U-Net%3A%20A%20Data-Driven%20Approach%20to%20Brain%20Tumor%20Segmentation%20Through%20Deep%20Learning%26lt%3B%5C%2Fspan%26gt%3B%26lt%3B%5C%2Fb%26gt%3B%26lt%3B%5C%2Fi%26gt%3B.%20%26lt%3Ba%20class%3D%26%23039%3Bzp-ItemURL%26%23039%3B%20href%3D%26%23039%3Bhttps%3A%5C%2F%5C%2Fwww.sciencedirect.com%5C%2Fscience%5C%2Farticle%5C%2Fpii%5C%2FS1746809425000011%26%23039%3B%26gt%3Bhttps%3A%5C%2F%5C%2Fwww.sciencedirect.com%5C%2Fscience%5C%2Farticle%5C%2Fpii%5C%2FS1746809425000011%26lt%3B%5C%2Fa%26gt%3B%26lt%3B%5C%2Fdiv%26gt%3B%5Cn%26lt%3B%5C%2Fdiv%26gt%3B%22%2C%22data%22%3A%7B%22itemType%22%3A%22document%22%2C%22title%22%3A%22BTS%20U-Net%3A%20A%20Data-Driven%20Approach%20to%20Brain%20Tumor%20Segmentation%20Through%20Deep%20Learning%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Carlos%22%2C%22lastName%22%3A%22Aumente-Maestro%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22David%20Rodr%5Cu00edguez%22%2C%22lastName%22%3A%22Gonz%5Cu00e1lez%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22David%22%2C%22lastName%22%3A%22Martinez%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Beatriz%22%2C%22lastName%22%3A%22Remeseiro%22%7D%5D%2C%22abstractNote%22%3A%22%22%2C%22date%22%3A%222025%22%2C%22language%22%3A%22%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fwww.sciencedirect.com%5C%2Fscience%5C%2Farticle%5C%2Fpii%5C%2FS1746809425000011%22%2C%22collections%22%3A%5B%22IYXGN3QT%22%2C%22IIGVPXZW%22%5D%2C%22dateModified%22%3A%222025-08-07T09%3A07%3A08Z%22%7D%7D%5D%7D
Aumente-Maestro, C., Müller, M., Remeseiro, B., Díez, J., & Reyes, M. (2025). AUDIT: An open-source Python library for AI model evaluation with use cases in MRI brain tumor segmentation. Computer Methods and Programs in Biomedicine, 108991. https://doi.org/10.1016/j.cmpb.2025.108991
Aumente-Maestro, C., Díez, J., & Remeseiro, B. (2025). A multi-task framework for breast cancer segmentation and classification in ultrasound imaging. Computer Methods and Programs in Biomedicine, 260, 108540. https://doi.org/10.1016/j.cmpb.2024.108540
Aumente-Maestro, C., González, D. R., Martinez, D., & Remeseiro, B. (2025). BTS U-Net: A Data-Driven Approach to Brain Tumor Segmentation Through Deep Learning. https://www.sciencedirect.com/science/article/pii/S1746809425000011
