In recent years, machine learning (ML) and artificial neural networks (ANNs), a particular subset of ML, have been adopted by various areas of healthcare. A number of diagnostic and prognostic algorithms have been designed and implemented across a range of orthopaedic sub-specialties to date, with many positive results. However, the methodology of many of these studies is flawed, and few compare the use of ML with the current approach in clinical practice. Spinal surgery has advanced rapidly over the past three decades, particularly in the areas of implant technology, advanced surgical techniques, biologics, and enhanced recovery protocols. It is therefore regarded an innovative field. Inevitably, spinal surgeons will wish to incorporate ML into their practice should models prove effective in diagnostic or prognostic terms. The purpose of this article is to review published studies that describe the application of neural networks to spinal surgery and which actively compare ANN models to contemporary clinical standards allowing evaluation of their efficacy, accuracy, and relatability. It also explores some of the limitations of the technology, which act to constrain the widespread adoption of neural networks for diagnostic and prognostic use in spinal care. Finally, it describes the necessary considerations should institutions wish to incorporate ANNs into their practices. In doing so, the aim of this review is to provide a practical approach for spinal surgeons to understand the relevant aspects of neural networks.

Cite this article: Bone Joint J 2021;103-B(9):1442–1448.

Aims

The aim of this study was to compare the clinical and radiological outcomes of patients with early-onset scoliosis (EOS), who had undergone spinal fusion after distraction-based spinal growth modulation using either traditional growing rods (TGRs) or magnetically controlled growing rods (MCGRs).

Aims

Psychoeducative prehabilitation to optimize surgical outcomes is relatively novel in spinal fusion surgery and, like most rehabilitation treatments, they are rarely well specified. Spinal fusion patients experience anxieties perioperatively about pain and immobility, which might prolong hospital length of stay (LOS). The aim of this prospective cohort study was to determine if a Preoperative Spinal Education (POSE) programme, specified using the Rehabilitation Treatment Specification System (RTSS) and designed to normalize expectations and reduce anxieties, was safe and reduced LOS.

Aims

To determine the value of scoliosis surgery, it is necessary to evaluate outcomes in domains that matter to patients. Since randomized trials on adolescent idiopathic scoliosis (AIS) are scarce, prospective cohort studies with comparable outcome measures are important. To enhance comparison, a core set of patient-related outcome measures is available. The aim of this study was to evaluate the outcomes of AIS fusion surgery at two-year follow-up using the core outcomes set.

There is a high incidence of spinal deformity in children with the Prader-Willi syndrome. We have encountered major complications following spinal surgery in this condition. We report our experience and conclude that spinal surgery is a formidable undertaking and the risks should be appreciated by the surgeon and the parents

The Matrix seating system is an adaptable orthosis made of interlocking plastic components which can be shaped to fit the needs of the disabled. Twenty-five patients who had used this system for a minimum of 12 months have been assessed clinically. It was found to have several advantages over its rivals particularly in patient and guardian acceptance, versatility and on economic grounds. However, it was found not to prevent deterioration in spinal deformity nor to prevent hip dislocation

A retrospective survey has been made of forty children with spinal tumours. Difficulties in establishing the correct diagnosis are mentioned and the value of radiological and cerebrospinal fluid investigations discussed. The major orthopaedic disabilities are spinal deformity or instability, and paraplegia. The main factor in the development of the former is the site of laminectomy: the higher the level the greater is the likelihood of deformity or instability developing. Measures to prevent this distressing complication are discussed. The role of the orthopaedic surgeon in the management of these children is emphasised

This retrospective study assesses the complications affecting the cervical spine after halo-pelvic traction in 83 patients who were followed up for a minimum of five years. Forty-four patients (53 per cent) had significant cervical complications such as radiological degenerative changes, avascular necrosis of the dens, loss of movement, pain or spontaneous fusion. The most important predisposing factors were a long period in the halo-pelvic apparatus, tuberculous kyphosis, stiffness of the spinal deformity and an age of 15 years or more at the time of application

Intramedullary spinal cord tumours may present as scoliosis without neurological signs. Those treating spinal deformities should be alert to this possible aetiology. The clinical features of 12 such cases are discussed with reference to early diagnosis and treatment. Patients with a painful scoliosis should be investigated with myelography as well as bone scintigraphy. Many intrinsic spinal cord tumours are now amenable to surgical removal. The prognosis for neurological recovery is poor once a severe deficit becomes established. The importance of early diagnosis and joint orthopaedic and neurosurgical management is emphasised

1. Two types of spinal extradural cyst are discussed: 1) the type which occurs in adolescents in the dorsal spine with evidence of kyphosis juvenilis; 2) the type which occurs in adults in the dorso-lumbar spine without kyphosis. 2. A fifth case of dorso-lumbar spinal extradural cyst in the adult without deformity is reported. Cysts in this region present common features distinguishing them from the more frequent type which occurs in adolescents. 3. The relationship between spinal extradural cysts, venous drainage of vertebral bodies, and spinal deformity, is discussed

1. The results of treatment have been compared in two unselected series of patients with unstable fractures of the thoraco-lumbar spine accompanied by paraplegia. 2. One group had been treated by conservative or " postural" methods while the others had been subjected to open reduction and internal fixation with double plates. 3. No difference in the amount of neurological recovery could be detected between the two groups but while a number of conservatively treated patients had significant residual spinal deformity and subsequently developed serious pain, this did not occur in any of the patients treated by plating. 4. It is concluded that open reduction and internal fixation are indicated in displaced fractures in the interests of long-term spinal function

We studied the possible role of melatonin deficiency in experimentally-induced scoliosis. A total of 90 chickens underwent pinealectomy on the third day after hatching: 30 were treated with serotonin, 30 with melatonin and 30 received no therapy (control group). Scoliosis developed in all the control group, in 22 of the serotonin group, and in only 6 of the melatonin group. The six melatonin-treated chickens with scoliosis had less severe spinal deformities than those in the serotonin-treated group. There were lower blood melatonin concentrations in chickens with scoliosis than in those without. Our findings suggest that melatonin deficiency contributes to the aetiology of this experimental scoliosis, probably by interfering with the normally symmetrical growth of the proprioceptive system involving the paraspinal muscles and the spine

Two hundred and fifty cases of myelodysplasia were reviewed in relation to spinal deformity. Approximately half of the children had, or were expected to develop, curves severe enough to need operations and only 10 per cent maintained completely undeformed spines. The most frequent deformity was scoliosis which could be subdivided into congenital and developmental types. The latter was of mixed aetiology, neuromuscular imbalance and asymmetry of the neural arch both contributing, while in some cases no causative factors could be identified. The best early indicator that developmental scoliosis was likely to appear was a high segmental level of both the neurological deficit and the neural arch defect. Deformity was very unlikely to start after the age of nine years

Routine examination for spinal deformity as part of a school health screening programme was introduced in Singapore in 1981. The three different ethnic groups included in the study provided figures for the prevalence of idiopathic scoliosis in an Asian population. A three-tier system of examination was used and a total of 110744 children in three age groups were studied. In those aged 6 to 7 years the prevalence was 0.12%. The prevalence in those aged 11 to 12 years was 1.7% for girls and 0.4% for boys, a ratio of 3.2 to 1. In girls aged 16 to 17 years the prevalence was 3.1%. In the latter two age groups there was a significantly higher prevalence in Chinese girls as compared with Malay and Indian girls. The optimal age for school screening seemed to be 11 to 12 years, but repeated examinations may be worthwhile

Thirteen patients, aged 7 to 45 years, have been treated for atypical forms of spinal tuberculosis at the Neurological Centres at Benghazi and Lahore. All presented with signs and symptoms of compression of the spinal cord or cauda equina, ranging from paraesthesiae and increasing weakness to paraplegia and loss of sphincter control. None of them showed visible or palpable spinal deformity nor the typical radiographic appearance of destruction of the intervertebral disc and the two adjoining vertebral bodies. These atypical cases fell into two well-defined groups: those with involvement of the neural arch only, with associated intraspinal cold abscesses, and those with involvement of a single vertebral body, resulting in its collapse and a radiographic appearance similar to that in secondary carcinoma of the vertebral body. The correct treatment in these two groups was diametrically opposed. Tuberculous disease of the neural arch was best traced by laminectomy; concertina collapse of a single vertebral body required cost-transversectomy and resection of the transverse process, the pedicle, and the portion of the vertebral body that was encroaching on the spinal canal

A radiological study of 50 patients with thoracic Scheuermann's disease revealed two types of lateral spinal curvature. A total of 43 lateral curves was present in 35 of the patients. Thirteen were apical at the same level as the Scheuermann's kyphosis and were due to vertebral-body wedging in the coronal plane; these curves had a mean Cobb angle of 15 degrees, occurred with equal prevalence in boys and girls and were directed equally to right and left. Thirty curves occurred in regions of compensatory lordosis (mean 5.6 degrees) situated above or, more commonly, below the Scheuermann's kyphosis. These scolioses had a mean Cobb angle of 16 degrees, were more often convex to the right than to the left and were significantly more prevalent in girls than in boys. The presence of these kyphoses and scolioses in the same spine, separated by only a few vertebrae, emphasises the importance of the sagittal plane in idiopathic spinal deformities and strongly suggests that idiopathic scoliosis and Scheuermann's disease share a common pathological process

Aims

The British Spine Registry (BSR) was introduced in May 2012 to be used as a web-based database for spinal surgeries carried out across the UK. Use of this database has been encouraged but not compulsory, which has led to a variable level of engagement in the UK. In 2019 NHS England and NHS Improvement introduced a new Best Practice Tariff (BPT) to encourage input of spinal surgical data on the BSR. The aim of our study was to assess the impact of the spinal BPT on compliance with the recording of surgical data on the BSR.

Since 1987, 22 children with myelomeningocele have been fitted with reciprocating orthoses. The level of the spinal lesions ranged from T10 to L4 and 13 had associated spinal deformities. Twelve of the patients currently use a Reciprocating Gait Orthosis, seven use a Hip Guidance Orthosis or Parawalker, one has progressed to a Knee Ankle Foot Orthosis, one has died and one has been lost to follow-up. The reciprocating orthoses are worn for a mean of 3.5 hours per day (1 to 6.5); daily usage by girls is almost twice that by boys. The mean daily usage by community walkers is 4.2 hours (13 children) as against 2.8 hours by household ambulators (8 children). Active hip flexion is not essential and fixed-flexion contractures up to 35 degrees can be accommodated. The average breakdown rate is 0.45 per year with an average of 1.5 adjustments each year. The average annual cost of a reciprocating orthosis is Aus$750 (375 pounds, US$570); this includes fabrication, adjustments and repairs