New gameplay of "oxygen therapy + AI"
From "manual adjustment" to "AI self-adaptation": the technical logic of the intelligent micro-compression oxygen chamber
The use of traditional micro-compression oxygen chambers depends on the experience of doctors or operators: patients need to inform their physical conditions in advance (such as whether they have high blood pressure or asthma), and doctors set the pressure (1.3-1.5ATA), oxygen concentration (90%-95%) and treatment time (usually 60 minutes) based on experience. The core breakthrough of the "AI intelligent micro-compression oxygen chamber" lies in the realization of the "real-time perception-analysis-intervention of hypoxia" closed loop through multi-sensor fusion and machine learning algorithms.
Its technical path can be divided into three steps:
Data collection: The cabin integrates a micro camera (capturing facial micro-expressions, such as cyanosis of the lips), an optical heart rate sensor (monitoring heart rate variability), and a blood oxygen saturation probe (real-time measurement of SpO₂), collecting 20 sets of physiological data per second;
AI analysis: Based on an algorithm model trained with millions of clinical hypoxia cases, it quickly determines the user's current hypoxia level (mild/moderate/severe), and associates age and medical history (health records entered by the user in advance) to generate a "risk profile";
Dynamic adjustment: According to the analysis results, the cabin automatically adjusts the pressure (for example, the pressure of moderate and severe hypoxia is increased to 1.5ATA), oxygen concentration (shortly increased to 98% in severe hypoxia), and continuously monitors during treatment. If an abnormal heart rate occurs (such as >100 beats/minute), the parameters are automatically reduced.
"Our algorithm can identify 'hidden hypoxia' that is difficult to capture with traditional equipment." The technical director of the technology company explained, "For example, some elderly people suffer from hypoxemia due to chronic obstructive pulmonary disease (COPD), but their blood oxygen saturation may only decrease slightly at rest. AI can provide early warning and adjust the plan by analyzing their respiratory rate and facial microcirculation changes."
Controversy 1: Can AI judgment replace medical blood gas analysis?
Although AI algorithms are promoted as "precise hypoxia judgment", the medical community remains cautious about its accuracy.
Liu Yang, chief physician of the Hyperbaric Oxygen Department of Peking Union Medical College Hospital, pointed out: "Medical blood gas analyzers can directly reflect the oxygen supply status of tissues by detecting indicators such as partial pressure of oxygen (PaO₂) and partial pressure of carbon dioxide (PaCO₂) in arterial blood. It is the 'gold standard' for clinical diagnosis of hypoxia. The AI algorithm relies on indirect physiological signals (such as heart rate and facial expressions), and the correlation between its data and the actual hypoxia state requires large-scale clinical verification."
He gave an example: "An anxiety patient may suffer from respiratory alkalosis (blood pH) due to hyperventilation. The value rises), at this time, AI may misjudge it as "hypoxia" and increase the oxygen concentration, which will increase the risk of alkalosis. AI cannot accurately distinguish such complex cases at present. "
The measured data of the hyperbaric oxygen center of a tertiary hospital also confirms this: in a comparative test of 100 patients with chronic hypoxia (COPD, sleep apnea), the AI algorithm's accuracy rate for "moderate hypoxia" was 78%, while the accuracy rate of medical blood gas analysis was 95%; the missed diagnosis rate of "hidden hypoxia" (hypoxemia without obvious symptoms) was 15%.
Controversy 2: Is the intelligent function an "efficiency-enhancing tool" or an "operation threshold"?
"The operation is too complicated!" Mr. Wang, a 72-year-old retired engineer, complained to reporters after using a certain brand of smart micro-pressure oxygen chamber. "The parameters on the screen jump too fast. I can't even tell the difference between 'pressure' and 'oxygen concentration'. I have to wait for the staff to adjust it every time."
Although the manufacturer claims that "AI makes oxygen therapy simpler", in actual experience, the "threshold" of smart functions is highlighted:
Data entry burden: Users need to enter information such as age, medical history, and commonly used drugs in advance (some devices require uploading physical examination reports), which is extremely unfriendly to elderly people with low cultural level or memory loss;
Difficulty in interpreting parameters: The screen in the cabin displays professional terms such as "hypoxia level" and "pressure trend chart" in real time. Most users said that "they don't understand it and only care about 'whether it is useful'";
Technology dependence risk: Once the device is disconnected from the network or the AI system fails (such as the camera is blocked), some smart models cannot manually adjust the parameters, which affects the safety of treatment.
A survey of a community nursing home in Shanghai showed that among elderly users over 65 years old, only 30% can independently start the "smart mode", and more than 50% still rely on staff to assist in operation. A staff member admitted: "Smart functions are convenient for young people, but have become a burden for the elderly."
Controversy 3: Can AI replace doctors? The clear answer from the medical community
Faced with the core question of "Can AI intelligent micro-pressure oxygen chamber replace doctors?", the medical community responded with a firm "No."
Professor Li Min, chairman of the Hyperbaric Oxygen Medicine Branch of the Chinese Medical Association, emphasized: "The essence of oxygen therapy is 'adjuvant treatment', and its plan needs to be formulated in combination with the patient's overall health status (such as underlying diseases, medication history, and allergy history). AI can provide data support, but the final decision-making power should be in the hands of doctors."
She further explained: "For example, a patient with atrial fibrillation who is taking anticoagulants needs to strictly control the pressure during oxygen therapy (to avoid barotrauma causing thrombosis); the duration of oxygen therapy for a pregnant woman needs to be strictly limited (high concentration of oxygen may affect the development of the fetal retina). In these complex situations, AI cannot make judgments based on physiological data alone, and must rely on the clinical experience of doctors."
Currently, no regulatory agency in any country in the world has approved the use of "AI micro-pressure oxygen chamber" as an "independent medical device." my country's "Safety Technical Specifications for Medical Microcompression Oxygen Chambers (Draft for Comments)" clearly stipulates: "AI functions can only be used as 'auxiliary tools', and the final parameter settings of the equipment must be confirmed by doctors; it is prohibited to promote expressions such as 'AI replaces doctors'."
Industry breakthrough: Let AI return to its "auxiliary" nature
Faced with controversy, leading companies have begun to adjust their direction. A technology company recently launched a "medical-grade intelligent microcompression oxygen chamber", while retaining the AI data monitoring function, added a "doctor-side APP" - doctors can view patients' physiological data through the cloud and adjust treatment plans remotely; the cabin cancels the complex screen, and only retains basic buttons such as "one-button start" and "emergency stop" to reduce the difficulty of operation for the elderly.
"Our goal is to 'let AI become the eyes of doctors, not replace doctors'." The person in charge of the company said frankly, "In the future, the value of intelligent microcompression oxygen chambers should be reflected in 'accurate recording', 'risk warning' and 'efficacy tracking', rather than replacing professional medical decisions."
Conclusion: AI is a tool, and health requires "human" warmth
From manual adjustment to AI self-adaptation, the intelligent road of microcompression oxygen chambers is essentially a deep integration of technology and health needs. It may improve efficiency and lower the threshold for operation, but it can never replace the clinical judgment and humanistic care of doctors.
As Chief Physician Liu Yang said: "No matter how advanced the technology is, the core of oxygen therapy is still 'people-oriented' - what we need is not 'thinking machines', but 'medical care that understands patients'." When AI learns to "assist" rather than "dominate", and when technology returns to the essence of "serving health", this exploration of "oxygen therapy + AI" can truly bring warm health solutions to the elderly and working people.
