Health & Research / AI news for Malaysia
From the archive · Event date 6 May 2025
UPM brought AI-assisted animal diagnosis into its veterinary hospital. What was proven?
The university said the system could detect disease in three minutes and support teaching, monitoring and research. Its public report did not include an accuracy study, sample size or patient outcomes.
In brief
- UPM's Faculty of Veterinary Medicine and veterinary technology company Aniware formalised a collaboration in Serdang on 6 May 2025.[1]
- UPM said the AI imaging system could help its veterinary hospital interpret medical data and detect disease in three minutes, while remaining a tool for doctors rather than a replacement.[1]
- The checked public sources do not provide a Malaysian validation study, diagnostic accuracy figures, sample size, clinical comparison or measured animal-health outcome.[1][2]
UPM opened a Malaysian clinical, teaching and research setting for Aniware's veterinary AI
Universiti Putra Malaysia's Faculty of Veterinary Medicine entered a strategic collaboration with Aniware Limited in Serdang on 6 May 2025. The university described the project as a way to use artificial intelligence in animal diagnostics and monitoring at Hospital Veterinar Universiti, while also opening work in education and research.[1]
The headline promise was speed. UPM quoted Dr Khor Kuan Hua, then head of its companion-animal medicine and surgery department, as saying Aniware's machine could detect disease in three minutes without waiting for a cardiologist's report. That is a consequential claim in urgent veterinary care, but the announcement did not publish the clinical evidence behind it.[1]
The most accurate reading is therefore narrower than a product breakthrough. UPM had created a pathway to use, teach and study the technology in a Malaysian veterinary setting. Whether it improved diagnostic accuracy, treatment decisions, waiting time or animal outcomes was still an open measurement question in the sources checked by Utopia.[1][2]
The collaboration covered clinical use, student learning and joint research
UPM said the system would automatically interpret complex medical data to help veterinary doctors make faster and more accurate decisions. The official photographs show CardioBird-branded equipment and an AI-Vitals display at the collaboration ceremony, indicating that cardiac or vital-sign monitoring formed part of the practical context presented that day.[1]
The partnership was not described as a single equipment purchase. UPM said it included AI-supported diagnostics and monitoring, development of educational courses related to animal disease, and opportunities for joint research. That gives the project three possible outputs: service at the hospital, exposure for students and evidence produced by researchers.[1]
Aniware's own website describes CardioBird X as an AI-powered system for vitals monitoring, electrocardiograms and Holter monitoring. It also advertises more than 240,000 datasets, operations in seven Asian regions and more than 1,000 animal-hospital partners. These are current company claims, not UPM findings or independently audited Malaysian results.[2]
A three-minute result matters only if clinicians know how reliable it is
A faster signal can help a veterinarian prioritise urgent cases, request further tests or contact a specialist sooner. It may also reduce routine interpretation work. But turnaround time and diagnostic quality are different measures: a result can be fast without being sufficiently sensitive, specific or clinically useful for every animal and condition.[1]
UPM's public report did not state how many animals were assessed, which diseases the system could detect, what reference standard was used or how its output compared with specialist interpretation. It also did not report false positives, false negatives, species coverage or whether the three-minute figure described processing time, total workflow time or a particular test.[1]
Those missing details do not mean the system failed. They mean readers should treat the speed figure as a university-published statement from the collaboration, not as proof of clinical superiority. A useful next publication would pair the operational claim with a local validation protocol and results that veterinarians can scrutinise.[1][2]
UPM positioned AI as decision support, not an autonomous animal doctor
Faculty dean Professor Goh Yong Meng said the system could help researchers record and track animal-patient data more efficiently. He also drew a clear boundary: the technology was intended to strengthen veterinarians with modern diagnostic tools, not replace them. That boundary is important when an AI output may influence treatment.[1]
For students, access inside a teaching hospital could make AI literacy concrete. Training should include how to read the system's confidence, recognise an out-of-distribution case, verify a result against clinical evidence, document overrides and explain uncertainty to an animal's owner. Ethical use is learned through these decisions, not through exposure to a device alone.[1]
The strongest proof would be a transparent outcome trail: cases processed, time saved, specialist agreement, corrections made by clinicians, changes in treatment and follow-up results. Publishing that evidence would help Malaysian veterinary practices distinguish a promising demonstration from a dependable clinical workflow.[1][2]
Why Malaysia should care
The collaboration places an AI-assisted diagnostic tool inside a Malaysian teaching hospital where clinicians, researchers and veterinary students can encounter it in real work. The value will depend on local validation, clinical oversight and transparent reporting rather than speed alone.
Veterinarians
The system may shorten an initial interpretation step, but the public announcement does not establish reliability across cases.[1]
Practical move: Use the output as decision support and record when clinical judgment confirms, corrects or rejects it.
Students and researchers
A teaching hospital creates access to real workflows and an opportunity to produce local evidence.[1]
Practical move: Study diagnostic performance, workflow time, human oversight and animal outcomes using a published protocol.
Animal owners
A faster result may support earlier decisions, but it should not be presented as a replacement for veterinary assessment.[1]
Practical move: Ask what the AI measured, how the veterinarian verified it and whether further tests or specialist review are needed.
What Malaysians can do now
- Publish the local validation design, sample size, supported species and conditions, and the reference standard used.
- Report diagnostic performance and workflow time separately, including false positives, false negatives and clinician overrides.
- Track whether AI-assisted decisions changed treatment, waiting time, referral patterns or animal-health outcomes.
What we still do not know
The collaboration and intended use are clear; the Malaysian performance evidence is not yet public.
- Which diseases, species and clinical signals were covered by the reported three-minute detection claim.
- How the system performed against specialist interpretation or another accepted reference standard at UPM.
- How many students, clinicians, researchers and animal cases later participated, and what outcomes resulted.
Sources
- 1.Kerjasama UPM dan Aniware Tingkat Kecekapan Diagnosis Haiwan dengan Teknologi AI Universiti Putra Malaysia, 7 May 2025
- 2.ANIWARE — Veterinary and Pet Artificial Intelligence Aniware Limited
